Small-molecule inhibitors of the MDM2-p53 protein-protein interaction based on an isoindolinone scaffold

Article abstract of DOI:10.1021/jm0601194

From a set of weakly potent lead compounds, using in silico screening and small library synthesis, a series of 2-alkyl-3-aryl-3-alkoxyisoindolinones has been identified as inhibitors of the MDM2-p53 interaction. Two of the most potent compounds, 2-benzyl-3-(4-chlorophenyl)-3-(3-hydroxypropoxy)-2,3- dihydroisoindol-1-one (76; IC₅₀ = 15.9 ± 0.8 μM) and 3-(4-chlorophenyl)-3-(4-hydroxy-3,5-dimethoxybenzyloxy)-2-propyl-2, 3-dihydroisoindol-1-one (79; IC₅₀ = 5.3 ± 0.9 μM), induced p53-dependent gene transcription, in a dose-dependent manner, in the MDM2 amplified, SJSA human sarcoma cell line.

Full text of DOI:10.1021/jm0601194

J. Med. Chem. 2006, 49, 6209-6221
6209
Small-Molecule Inhibitors of the MDM2-p53 Protein-Protein Interaction Based on an
Isoindolinone Scaffold
Ian R. Hardcastle,*,† Shafiq U. Ahmed,‡ Helen Atkins,‡ Gillian Farnie,‡ Bernard T. Golding,† Roger J. Griffin,†
Sabrina Guyenne,† Claire Hutton,‡ Per Ka¨llblad,§ Stuart J. Kemp,† Martin S. Kitching,† David R. Newell,‡ Stefano Norbedo,†
Julian S. Northen,† Rebecca J. Reid,† K. Saravanan,† Henrie¨tte M. G. Willems,*,§ and John Lunec*,‡
Northern Institute for Cancer Research, School of Natural SciencessChemistry, Bedson Building, UniVersity of Newcastle upon Tyne,
Newcastle, NE1 7RU, United Kingdom, Northern Institute for Cancer Research, Paul O’Gorman Building, Medical School, Framlington Place,
UniVersity of Newcastle upon Tyne, Newcastle, NE2 4HH, United Kingdom, and De NoVo Pharmaceuticals, Compass House, Vision Park,
Histon, Cambridge, CB4 9ZR, United Kingdom
ReceiVed February 3, 2006
From a set of weakly potent lead compounds, using in silico screening and small library synthesis, a series
of 2-alkyl-3-aryl-3-alkoxyisoindolinones has been identified as inhibitors of the MDM2-p53 interaction.
Two of the most potent compounds, 2-benzyl-3-(4-chlorophenyl)-3-(3-hydroxypropoxy)-2,3-dihydroisoindol-
1-one (76; IC50 ) 15.9 ( 0.8 µM) and 3-(4-chlorophenyl)-3-(4-hydroxy-3,5-dimethoxybenzyloxy)-2-propyl-
2,3-dihydroisoindol-1-one (79; IC50 ) 5.3 ( 0.9 µM), induced p53-dependent gene transcription, in a dose-
dependent manner, in the MDM2 amplified, SJSA human sarcoma cell line.
Introduction
The X-ray crystal structure of the N-terminal domain of
MDM2 bound to a 15-residue peptide, corresponding to a region
of the transactivation domain of p53, which had been mapped
as the binding site for MDM2, revealed a relatively small
binding pocket on MDM2.16 The protein interacts with three
amino acid residues within a hexapeptide region (T18 F S D L
W23) of p53 arrayed in an amphipathic R-helix. The tractability
of this interaction as a target for small molecule inhibitors was
suggested by peptide mapping experiments, which defined a
hexapeptide region of p53 as the consensus sequence for MDM2
binding.16 Phage-display peptide libraries were used to identify
the 8-mer IP3 peptide (Ac-Phe19-Met-Asp-Tyr-Trp-Glu-Gly-
Leu26-NH2) that inhibits p53 binding to MDM2 (IC50 ) 9 µM).17
This sequence was used as a lead for the rational design of the
potent AP peptide (Ac-Phe19-Met-Aib-Pmp-6-Cl-Trp-Glu-Ac3c-
Leu26-NH2; IC50 ) 5 nM) in which replacements with unnatural
amino acids have been used to stabilize the R-helical structure
and to introduce a salt-bridge interaction between the Pmp
residue and Lys94 outside the p53 binding domain of MDM2.18
A number of small-molecule inhibitors of the p53-MDM2
interaction have been reported, for example, chalcone deriva-
tives, such as 1, which are weak inhibitors (IC50 ) 49 µM).19
Computational methods have been used to identify potential
inhibitors from combinatorial scaffolds, for example, 2, some
of which have been shown to display inhibitory activity in cell-
free and cellular assays.20,21 The natural product chlorofusin
inhibits MDM2-p53 binding with an IC50 of 4.6 µM.22 More
recently, highly potent small-molecule inhibitors have been
reported, including the Nutlins (e.g., 3a,b),20,22-24 and benzo-
diazepinediones, such as 4.25,26
The p53 tumor suppressor acts as “the guardian of the
genome” by reacting to cellular stress, which may be caused
by hypoxia and DNA damage. This protein activates the
transcription of a number of genes that govern progression
through the cell cycle, the initiation of DNA repair, and
programmed cell death.1,2 The activity of p53 is tightly regulated
by the MDM2 protein, which is itself transcriptionally trans-
activated by p53. MDM2 binds to and inactivates the p53
transactivation domain and also ubiquitylates the MDM2-p53
complex to target it for proteosomal degradation. In normal cells,
the balance between active p53 and inactive MDM2-bound p53
is maintained in a negative feedback loop.3,4
Inactivation of p53 through mutation is common in ap-
proximately a half of all tumors. In about 7% of tumors,
amplification and overexpression of the MDM2 gene results in
the loss of functional p53, allowing transformation and uncon-
trolled tumor growth.5 Inhibitors of the MDM2-p53 binding
interaction are expected to restore normal p53 activity in
MDM2-overexpressing cells and thus exert an antitumor effect.6
Traditionally, the disruption of protein-protein interactions
has been regarded as a difficult or even impossible goal for
drug development. These interactions occur over large surface
areas (average ∼ 800 Å2), frequently with relatively flat
topologies, typically have high binding affinities resulting from
the sum of a large number of relatively weak noncovalent bonds.
This suggests that each protein will be devoid of high-affinity
sites for small-molecule inhibitors.7,8 However, several examples
of small-molecule antagonists of protein-receptor targets have
been reported recently.9,10 The targets include: IL-2,11 BclXL
/
Bax interactions,12 p60c-src SH2 interactions with phosphoryl-
ated tyrosine residues,13 the association of HIV-1 Tat with
PCAF,14 and bid/Bax interactions.15
The cellular effects of MDM2-p53 inhibition have been
investigated using a variety of agents. Inhibition of MDM2
expression with antisense oligonucleotides in cell lines with
elevated and normal levels of MDM2, for example, JAR, SJSA,
and MCF7 cell lines, was observed to induce increased levels
of p53.27,28 Additionally, apoptosis was observed with the JAR
cells. Antisense oligonucleotides synergistically increased the
potency of the DNA damaging agent camptothecin in MDM2-
overexpressing and normal cell lines.29 Peptide inhibitors have
also been used to evaluate the effects of MDM2-p53 inhibition.
Wasylyk et al. observed a cytotoxic effect when the IP3 peptide
* Corresponding authors: Tel.: +44 191 222 6645 (I.R.H.); +44 191
246 4420 (J.L.); +44 1223 238012 (H.W.). Fax: +44 191 8591 (I.R.H.);
+44 191 4301 (J.L.); +44 1223 238088 (H.W.). E-mail: i.r.hardcastle@
ncl.ac.uk (I.R.H.); john.lunec@ncl.ac.uk (J.L.); henriette.willems@
denovopharma.com (H.W.).
† School of Natural SciencessChemistry, University of Newcastle upon
Tyne.
‡ Medical School, University of Newcastle upon Tyne.
§ De Novo Pharmaceuticals.
10.1021/jm0601194 CCC: $33.50 © 2006 American Chemical Society
Published on Web 09/28/2006

6210 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 21
Hardcastle et al.
induce cellular levels of p53, MDM2, and p21Waf1/Cip1 in
the p53 wild-type HCT116 cell line, but not in the p53 mutant
SW480. This small-molecule antagonist also showed a dose-
dependent antiproliferative activity, which was dependent on
the p53 status of the cell line. An in vivo antitumor effect was
also demonstrated for Nutlin-3 in an MDM2 amplified SJSA-1
xenograft model at a 10 mg/kg dose.24
In this paper, we present a full report of the development of
small-molecule inhibitors of the MDM2-p53 interaction based
on an isoindolinone scaffold, which display cellular activity.32
Preliminary screening studies, using an in vitro p53-MDM2
binding assay, identified compounds 5, 6, and 7 as weak
inhibitors of the p53-MDM2 interaction (IC50 ∼ 200 µM).
These compounds also displayed growth inhibitory activity in
the NCI 60 cell line screen, and importantly, were rated
COMPARE negative with respect to any known classes of
antitumor agents.33 The versatility of the isoindolinone scaffold
for the synthesis of combinatorial libraries and the availability
of a high-resolution X-ray structure of the MDM2 protein target
in complex with a p53-derived peptide encouraged us to pursue
the optimization of this class of compounds. A program of a
focused library synthesis, guided by virtual screening, has
resulted in the discovery of novel isoindolinone inhibitors of
the MDM2-p53 interaction with low micromolar affinity and
cellular activity.
was expressed in cells.30 The more potent AP peptide, used
exogenously, has been shown to induce p53, albeit weakly, in
HCT116, JAR, and SJSA-1 cell lines, which express low,
intermediate, and high MDM2 levels, respectively. A large
induction of the p21Waf1/Cip1 cyclin-dependent kinase inhibitor,
the gene for which is a transcriptional target of p53, was also
observed in all three cell lines, but not in p53 null and p53
mutant cell lines. Induction of apoptosis was observed in the
JAR and SJSA-1 cell lines.31 Nutlin-3 was demonstrated to
Computational Chemistry
In the absence of structural data on the isoindolinone-MDM2
interaction, docking was applied to determine a single, low-
energy binding mode for the initial lead compounds 5 and 7
Table 1. Inhibition of the MDM2-p53 Binding Interaction by Seed Compounds Used in the Second Round of Binding-Mode Determinations

Isoindolinone MDM2-p53 Inhibitors
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 21 6211
Scheme 1. Synthesis of Substituted Isoindolinones from the ψ-Acid Chloride 9a
a Reagents and conditions: (a) SOCl2, DMF, THF; (b) R1NH2, Et3N, THF; (c) SOCl2, DMF, THF; (d) R2OH or R2NH2, K2CO3 or Et3N, THF.
Scheme 2. Synthesis of Substituted Isoindolinones from the
Thioethers 12a and 12ba
round of virtual screening of new substituents. A selection of
these “virtual hits” were synthesized. Disappointingly, the
compounds suggested by the virtual screening described above
were no more active than a series of compounds selected for
synthesis on the basis of diversity and accessibility (see
Supporting Information).
The strategy for the binding mode prediction was reassessed
in the light of these preliminary results. The open and lipophilic
nature of the p53 binding site on MDM2 makes the determi-
nation of a single binding mode for the isoindolinone scaffold
difficult. The possibility that there are multiple binding modes
for the scaffold, depending on the substituents present, has to
be considered. The use of multiple binding modes in this setting
has been shown to increase significantly the probability of
predicting correctly the experimental binding mode.35 This
approach was deemed more likely to succeed with the selection
of a larger, more diverse set of “seed” compounds in the binding
mode determination.
a Reagents and conditions: (a) SOCl2, DMF, THF; (b) PhCH2SH, Et3N,
THF; (c) R2OH, NIS, CSA, THF.
Scheme 3. Synthesis of Substituted Isoindolinone 53a
An ensemble of six “seed” compounds were used, with an
IC50 range from 27 to 92 µM (8-13; Table 1), as the basis for
a second round of virtual screening. Each of these compounds
was docked into the MDM2 crystal structure 1YCR,16 using
the programs easyDock34 and GOLD,36 again resulting in a large
number of solutions. A total of 24 high-scoring, unique binding
modes were chosen providing starting points for the second
round of virtual screening, that is, one binding mode per
compound, stereoisomer, and docking program combination.
For each binding mode, the docked position of the isoindoli-
none scaffold was preserved and novel ligands were generated
by the addition of either R1 or R2 substituents from a list of
commercially available reagents. The interaction between the
ligands and the MDM2 protein was explored using a simulated
annealing optimization of an empirical scoring function using
the program Skelgen.37 Ligands able to form at least one H-bond
a Reagents and conditions: (a) (i) SOCl2, DMF, THF; (ii) n-PrOH, THF;
(b) TBAF, THF.
with MDM2. Both stereoisomers of these compounds were
docked into the published MDM2 crystal structure 1YCR,16
using the program easyDock.34 As a result of the lipophilic
properties of the compounds and the open shape and lipophilicity
of the MDM2 binding site, a large number of possible docking
solutions were identified. A single, high-scoring binding mode
was chosen for both 5 and 7 from a large number of low-energy
docking solutions. This was selected as the basis for the first
Scheme 4. Synthesis of Substituted Isoindolinone 57a
a Reagents and conditions: (a) SOCl2, DMF, THF; (b) TBDMSOC4H8NH2, THF; (c) SOCl2, DMF, THF; (d) TBAF, THF; (e) COCl2, DMSO, -78 °C;
(f) MeOH, NH4Cl.

6212 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 21
Scheme 5. Synthesis of Substituted Isoindolinones 24 and 30a
Hardcastle et al.
Scheme 8. Synthesis of Substituted Isoindolinone 47a
a Reagents and conditions: (a) (i) SOCl2, DMF, THF; (ii) (R)-2-
(TBDMSO)-2-(4-methoxyphenyl)ethylamine, Et3N, DMF; (b) TBAF, THF.
a Reagents and conditions: (a) (i) SOCl2, DMF, THF; (ii) 4-allyloxy-
benzyl alcohol or 3-allyloxybenzyl alcohol, K2CO3, THF; (b) K2CO3,
Pd(PPh3)4, MeOH.
Scheme 9. Synthesis of Substituted Isoindolinone 65a
Scheme 6. Synthesis of Substituted Isoindolinone 67a
a Reagents and conditions: (a) (i) SOCl2, DMF, THF; (b) 2-aminoethyl
methacrylate hydrochloride, Et3N, THF; (b) SOCl2, DMF, THF; (c)
2-(TBDMSO)-1-(TBDMSOmethyl)-2-phenylethylamine, Et3N, DMF; (d)
TBAF, THF.
a Reagents and conditions: (a) ethylenediamine, THF; (b) Ac2O, pyridine;
(c) SOCl2, DMF, THF; (d) 4-tert-butylbenzyl alcohol, Et3N, THF.
Scheme 7. Synthesis of Substituted Isoindolinone 43a
Scheme 10. Synthesis of Substituted Isoindolinones 28 and 41a
a Reagents and conditions: (a) (i) SOCl2, DMF, THF; (ii) (R)-2-
(TBDMSO)-2-(4-methoxyphenyl)ethylamine, Et3N, DMF; (b) TBAF, THF.
with target residues of MDM2 were selected as “virtual hits”
and suggested for synthesis, and a set of 57 ligands were
identified and suggested for synthesis.
Chemistry
Substituted isoindolinones were prepared according to Scheme
1. The appropriate benzoylbenzoic acids (8) were converted into
the ψ-acid chlorides (9) under Vilsmeier conditions and
condensed with the R1-primary amines to give the 4-hydroxy-
isoindolinones (10). The 4-hydroxy compounds (10) were
converted into their respective chlorides (11) and, subsequently,
reacted with appropriate alcohols in the presence of base (Et3N
or K2CO3) to give the final substituted isoindolinones (13, 19-
22, 25, 26, 31-40, 44, 45, 48-51, 64, 65, 68-73, 75, and 78-
83) as racemates (Scheme 1).38
a Reagents and conditions: (a) 2-bromoethanol, THF; (b) piperazine,
MeOH, reflux; (c) sodium sulfite, DME, water, reflux.
to nucleophilic displacement by oxidation is an established
strategy in the synthesis of substituted pyrimidines, with
m-chloroperbenzoic acid as oxidant.39 To investigate this route,
4-hydroxyl compounds 10 were converted into the unstable
chlorides (11) and trapped with benzylmercaptan to give the
stable thioethers (12a,b). The oxidation of 12a or 12b using
mCPBA was attempted, followed by displacement with an
alcohol. However, this resulted in decomposition of the starting
material, with no product identified by LC-MS or NMR in the
crude reaction mixture. Successful activation of the thioethers
The instability of the intermediate chlorides (11) to hydrolysis
posed some difficulties in their efficient handling in a parallel
synthesis approach. Consequently, we sought a stable intermedi-
ate that could be activated in situ. The activation of thioethers

Isoindolinone MDM2-p53 Inhibitors
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 21 6213
Scheme 11. Synthesis of Substituted Isoindolinones via a
Directed ortho-Metalation Reactiona
was required for both the 2-(3-hydroxypropyl) substituent and
the secondary hydroxyl group of the 2-hydroxy-2-(4-methoxy-
phenyl)ethanol used in the synthesis of 43 (Scheme 7). Silyl
protection was also employed in the synthesis of 47 from
2-hydroxy-(4-methoxyphenyl)ethanol (Scheme 8).
The synthesis of 2-ethylacetamide derivatives could be
achieved by direct synthesis, albeit in low yield. In the case of
the tert-butylbenzyloxy derivative, ethane-1,2-diamine was
reacted with the ψ-acid chloride to give the isoindolinone 10u,
which was acetylated to 10v prior to conversion into the chloride
and subsequent reaction to give 65 (Scheme 9).
a Reagents and conditions: (a) s-BuLi, TMEDA, THF, -78 °C; (b) (i)
SOCl2, DMF, THF; (ii) R2OH, Et3N, THF; (c) RCOCl, Et3N, DCM.
Two isoindolinones were synthesized from the 2-bromoethoxy
isoindolinone (27) as the common intermediate. Nucleophilic
displacement of bromide with either piperazine or sulfite gave
28 and 41, respectively (Scheme 10).
12a and 12b to nucleophilic displacement was achieved by
treatment with N-iodosuccinimide (NIS) in the presence of
catalytic camphorsulfonic acid (CSA).40 The oxidized intermedi-
ates were reacted in situ with the appropriate R2-alcohol to give
compounds 74, 76, and 77 (Scheme 2).
An alternative route to substituted 2-n-propyl isoindolinones
proceeded by directed ortho-metalation of n-propylbenzamide
and reaction with the appropriate benzoate ester or benzonitrile
to give the hydroxyisoindolinone 10l or aminoisoindolinones
10m,n, respectively. The hydroxyisoindolinone 10l was con-
verted into the chloride and substituted with the appropriate
alcohol or amine to give the desired isoindolinones 58-64 and
84-87. Aminoisoindolinones 10m,n were treated with the
appropriate acid chloride to give the amides 16 and 17,
respectively (Scheme 11).
The silylethoxymethoxybenzoic acid (92) was prepared from
phenolphthalein by hydrolysis to the benzoic acid (88),41 which
was esterified to 89 before conversion to the SEM ether 90 and
saponification to 91. The benzoic acid 91 was converted into
the ψ-acid chloride and onto the hydroxyisoindolinone (10i).
Isoindolinone 10i was converted into the chloride and treated
with syringic alcohol to give 92 (Scheme 12).
In selected cases, where the R1-amine substituent contained
additional nucleophilic amino or hydroxyl groups, a protecting
group strategy was required. Similarly, isoindolinones 10q,r
were prepared using the required tert-butyldimethylsilyl-
protected amino alcohols, which proved stable under the
Vilsmeier conditions employed. Following appropriate substitu-
tion at C-3 to give 52 and 54, deprotection with tetrabutyl-
ammonium fluoride (TBAF) gave the alcohols 53 and 55,
respectively (Schemes 3 and 4). The alcohol 55 was used as
the precursor to the dimethylacetal 57. Oxidation of 55 under
Swern conditions gave the aldehyde 56, which was converted
into the 57 under mild acidic conditions in methanol.
Protection was also required to ensure the correct regio-
selectivity for R2-substitutions with multiple nucleophilic groups.
The 3- and 4-hydroxybenzyl alcohol substituents were intro-
duced as their allyl ethers giving 29 and 23, respectively.
Deprotection was effected under Pd(0) catalysis with methanol
and potassium carbonate giving phenols 30 and 24, respectively
(Scheme 5). 2-Hydroxy-1-hydroxymethyl-2-phenylethylamine
was introduced as its bis-tert-butyldimethylsilyl ether giving 67
following deprotection (Scheme 6). Similarly, silyl protection
Results and Discussion
Structure-Activity Relationships for Inhibition of the
MDM2-p53 Interaction. The virtual screening using multiple
binding modes suggested a large set of novel ligands. A set of
Scheme 12. Synthesis of Substituted Isoindolinone 92a
a Reagents and conditions: (a) KOH, hydroxylamine hydrochloride, H2O, 80 °C; (b) AcCl, MeOH; (c) Cs2CO3, SEMCl, CH3CN; (d) TMSOK, DCM; (e)
SOCl2, DMF, THF; (f) BnNH2, Et3N, THF; (g) SOCl2, DMF, THF; (h) syringic alcohol, THF.

6214 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 21
Hardcastle et al.
Table 2. Inhibition of the MDM2-p53 Binding Interaction by Isoindolinones with R1 and R2 Reagents Selected from the Second Virtual Screen,
Considering Multiple Binding Modes

Isoindolinone MDM2-p53 Inhibitors
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 21 6215
Table 2 (Continued)
43 compounds from this set were prepared, covering 14 of the
24 binding modes (BM01-BM24; Table 2). Where possible,
substituents that only appeared in a single binding mode were
chosen with the intention that a small number of preferred
binding modes would be identified. Compounds with such
substituents are labeled unique in Table 2. Of the 43 compounds
evaluated, five displayed little or no inhibitory activity (21, 44,
45, 48, and 53; IC50 > 400 µM). Approximately half of the
group inhibited in the range 100-400 µM. Twelve compounds
(22, 24, 30, 34, 35, 49, 57, and 59-63) displayed similar activity
(IC50 ) 50-100 µM) to the six seed compounds used in the
binding mode study. Two compounds (40 and 65) displayed
improved activity over the most potent seed compound 16.
Compounds displaying similar or improved activity were
identified from a number of potential binding modes, suggesting
that the isoindolinone may adopt different orientations within
the MDM2 binding pocket depending on the R1 and R2
substituents. A number of binding modes suggested compounds
with no significant activity (e.g., BM01, BM03, and BM11),
and so, it may be reasonable to disregard these orientations. In
contrast, two binding modes, BM06 and BM10, are of interest,
as they suggested the most active compounds 40 and 65,
respectively (Figure 1).
BM06 models the isoindolinone scaffold occupying the Trp23
binding pocket, the N-ethylacetamide group making a hydrogen
bond to Gln72 in the region occupied by Phe19 of p53 and the
tert-butylbenzyl group overlaying the surface of MDM2. In
contrast, BM10 models the isoindolinone scaffold overlaying
the surface of MDM2 with the Trp23 pocket occupied by the
3-phenyl group, the syringic alcohol substituent occupies the
Leu26 pocket and makes a hydrogen bond to Tyr100, and the
N-benzyl substituent occupies the Phe19 pocket. In comparison,
the X-ray structures of Nutlin-2 (3a) and the 4, both the Leu26
and Trp23 pockets are occupied by 4-bromophenyl and 4-chloro-
phenyl residues, respectively, and the Phe19 pocket is occupied
by a combination of the 2-ethoxy-4-methoxyphenyl and the
N-hydroxyethylpiperazine residues for 3a and the 7-iodobenzo-
diazepinedione ring of 4.
Bearing in mind that these compounds were uniquely
identified for specific binding modes it is tempting to consider
these binding modes as possible experimental solutions. The
SARs indicate that multiple binding modes are likely to exist

6216 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 21
Hardcastle et al.
Figure 1. Compounds 65 and 40 docked into MDM2 structure (1YCR) showing binding modes BM06 and BM10, respectively, and cartoons
indicating the predicted hydrogen bond; p53 residues (purple) are included for reference.
for the isoindolinone compounds. The fact that binding modes
BM06 and BM10 also suggested a number of compounds with
poorer activity than the seed compounds might not necessarily
indicate a flaw in the binding mode prediction, but perhaps
instead that those reagents were false positives in the virtual
screen. Determination of the structure of the isoindolinone-
MDM2 complex is in progress and will provide experimental
validation of the binding modes and shed further light onto the
results of the virtual screening.
In the absence of a confirmed binding mode for the iso-
indolinone inhibitors, we decided to employ a combinatorial
approach to optimize the activity of this series of compounds.
Based on the experimental data (Tables 1-2), the most favorable
substituents at each of the three positions of variation on the
isoindolinone template were identified. At the R1 position
benzyl, n-propyl, and 2-acetamidoethyl were selected. At the
R2 position phenyl, 4-chlorophenyl, and 4-trimethylsilyl-
ethoxymethoxyphenyl (4-SEMOPh) were chosen as the most
favorable aromatic substituents, and 4-tert-butylbenzoxy, 3,5-
dimethoxy-4-hydroxybenzoxy, 3-hydroxypropoxy, and 2-(2-
pyridyl)ethoxy were selected as the ether substituents conferring
the best activity. An array of 36 isoindolinones of covering every
possible combination of substituent was designed. For certain
substituents, where the preliminary biological activity was
discouraging and the synthesis lengthy or problematic, the
synthesis of every combination was not attempted. A total of
24 array compounds were synthesized and assayed (Table 3).
The results shown in Table 3 indicate the following: 68-
70, 72, 73, 78, 84-87, and 92 displayed lower potency than
their parents; 13, 71, 74, 75, and 81-83 displayed comparable
activity with the parent compounds; and 40, 65, 76, 77, 79, and
80 were the most potent in the group. A comparison of the
results between the R2 substituents reveals that the 4-trimethyl-
silylethoxymethoxyphenyl was not favorable in all cases. The
modest activity of some compounds bearing the 4-SEMOPh may
be due to nonspecific hydrophobic interactions between the
bulky lipophilic substituent and the largely hydrophobic binding
pocket of MDM2. Comparison between the R2-phenyl and
4-chlorophenyl series reveals contrasting SAR for the N-benzyl
and N-n-propyl substituents. In the N-n-propyl series, the
4-chlorophenyl substituent confers improved activity over the
parent phenyl in every case. The effect is most striking between
the 3-hydroxypropoxy derivative 72, which is essentially
inactive, and 80 (IC50 ) 16.4 µM). In the case of the syringic
alcohol derivatives 71 and 79, a 16-fold gain in activity is
observed for the 4-chloro compound. In contrast, in the N-benzyl
series, the 4-chloro substituent is favorable in the case of the
3-hydroxypropoxy and 2-(2-pyridyl)ethoxy derivatives (68 and
69 vs 76 and 77) but unfavorable for the syringic alcohol
substituent (40 vs 75). For the one comparable example in the
N-(2-acetamidoethyl) series the 4-chloro substituent is not
favored. These results suggest that the N-benzyl and N-n-propyl
isoindolinones bind to MDM2 in different orientations, resulting
in different SARs. The most potent isoindolinone identified from
the array is 79 (NU8231; IC50 ) 5.3 ( 0.9), which shows a
4-fold improvement in potency over the parent 40.
The presence of 4-haloaromatic substituents in potent MDM2-
p53 inhibitors demonstrates that these substituents produce
favorable binding interactions in both the Trp23 and the Leu26
binding pockets.23-26 The gains in potency seen for the 4-chloro-

Isoindolinone MDM2-p53 Inhibitors
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 21 6217
Table 3. Inhibition of the MDM2-p53 Binding Interaction by
Isoindolinones Members of the Combinatorial Array
Figure 2. Western blot analysis of SJSA cells treated with increasing
concentrations (5-20 µM) of compounds 76 and 79, showing the
induction of dose-dependent increases in MDM2 and p21 protein levels.
Figure 3. Change in p53-dependent luciferase activity in SJSA cells
treated with increasing concentrations (5-40 µM) of 76 for 6 h. The
bars represent the mean and standard errors for three independent
experiments. Increases that were significantly higher (p < 0.05; two-
tailed t-test) than the DMSO control level (broken line) are marked by
an asterisk.
Cell Biology. The most potent compounds identified were
selected for further evaluation to ascertain whether they had
effects on intact cells that were consistent with inhibition of
MDM2-p53 binding. Compounds 76 and 79 showed significant
biological activity. SJSA cells, in which the MDM2 gene is
amplified, were treated with increasing concentrations (5, 10,
and 20 µM) of 76 and 79. Cells were lysed at 6 h and protein
extracts were analyzed by western immunoblotting for p53,
p21WAF1, and actin (Figure 1). Compounds 76 and 79 produced
a dose-dependent increase in MDM2 and p21, consistent with
the release of p53 transcriptional activity. Surprisingly, the
accumulation of p53 protein levels was not observed in the SJSA
cells under these conditions. To test for further evidence of
increased p53 transcriptional activity in response to treatment
with the compounds, a luciferase-based reporter gene assay was
used. This assay again showed the induction of a dose-dependent
increase in p53 transcriptional activity for both 76 (Figures 2
and 3) and 79 (data not shown).
Conclusions
We have demonstrated the utility of docking and in silico
combinatorial methods in the optimization of a series of
inhibitors of the MDM2-p53 interaction based on an iso-
indolinone scaffold. We consider that the isoindolinones pre-
sented are an interesting addition to the small-molecule inhibitors
of the MDM2-p53 protein-protein interaction. Further opti-
mization of this class of compounds is in progress.
phenyl-substituted isoindolinones series are consistent with these
observations and with the predicted binding mode BM10.

6218 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 21
Hardcastle et al.
when the appropriate constraints are satisfied and conferred a
positive value otherwise. The constraints used in this work included
intramolecular ligand van der Waals clashes and intermolecular
ligand-protein van der Waals clashes, both according to the Bondi
set.46 A reagent was deemed to be successful in the virtual screen,
yielding a “hit”, if it was able to form at least one H-bond with the
following H-bond anchoring points of MDM2: Leu 54 O, Phe 55
O, Gln 59 NE2, Gln 72 O and OE1, Val 93 O, His 96 ND1, and
Tyr 100 OH. The isoindolinone scaffold was held rigid during the
screen. For ranking purposes, a scoring function very similar to
Bo¨hm’s empirical scoring function was used as a measure of
ligand-protein interactions.47,48 The scoring function is knowledge-
based and includes H bonds, ionic interactions, lipophilic protein-
ligand contacts, and the number of rotatable bonds in the ligand.
Synthesis. General Procedure A: Synthesis of 3-Hydroxy-
isoindolin-1-ones. To a solution of the appropriate 2-benzoyl-
benzoic acid (1.0 equiv) in THF was added thionyl chloride (2.2
equiv) and DMF (3 drops). The mixture was stirred at room
temperature 16 h and then concentrated in vacuo to give a clear
oil. The residues were dissolved in THF (10 mL), the appropriate
primary amine (1.0 equiv) and triethylamine (2.2 equiv.) were
added, and the mixture was stirred at room temperature for 16 h.
The mixture was either filtered and submitted to extraction with
EtOAc (15 mL), sodium bicarbonate (20 mL), and water (15 mL)
or treated immediately with EtOAc (15 mL), saturated sodium
bicarbonate (15 mL), and water (15 mL). The organic layers were
combined, dried (Na2SO4), and concentrated in vacuo. Chroma-
tography (EtOAc, petrol 1:4) or crystallization, with a minimum
of EtOAc and an excess of petrol, gave the desired product.
General Procedure B: 3-Alkoxy-2,3-dihydroisoindol-1-ones.
To a solution of 11a (0.51 g, 1.50 mmol) in THF (7 mL) was added
the appropriate alcohol (3.0 or 4.0 mol equiv). The reactions were
stirred at room temperature for 72 h, unless stated otherwise, and
then concentrated in vacuo. The residue was dissolved in EtOAc
(20 mL) and washed with water (3 × 10 mL). The organic layer
was dried (MgSO4) and concentrated in vacuo to give the crude
3-alkoxy-2,3-dihydroisoindol-1-one.
Experimental Section
Computational Chemistry. The computational methods used
for this work include binding-mode prediction through docking and
virtual screening of substituents for the isoindolinone scaffold. For
both purposes, the crystal structure 1YCR16 of MDM2 was prepared
by removing the p53 peptide and adding hydrogen atoms in
InsightII/Discover3 (Accelrys, Inc., San Diego, CA) at pH 7.0. Their
positions were subject to 500 iterations of steepest descent
minimization, followed by 500 steps of conjugate gradient mini-
mization using the CFF force field.42 All nonhydrogen atoms were
held rigid during these minimizations.
Docking studies were carried out using the programs easyDock34
and GOLD.36 The selection of binding modes from the large pool
of solutions for the second round of docking was performed
according to a protocol described previously.35 In brief, the ligand
binding modes were ordered according to their binding energies.
Then, starting from the lowest-energy solution, the rmsd of each
pose with respect to all the previous lower-energy poses was
computed. If the rmsd was higher than a given threshold (1.5 Å),
then that pose was added to a list of distinct solutions. The result
of this procedure was a list of distinct ligand-binding mode poses
that had rmsds between one another of more than the predefined
threshold. The number of binding modes selected to be used in the
second round of virtual screening was set to 24. This number was
based on the fact that six compounds were used for the docking
study, each of these with one stereocenter on the isoindolinone
scaffold, and that two different docking programs were used.
Reagents derived from primary amines (R1), acid chlorides (R2),
and primary alcohols (R2) were extracted from the Daylight
Available Chemicals Directory (ACD; Daylight Chemical Informa-
tion Systems, Inc., CA). Reagents containing atoms other than C,
O, N, S, Na, Ca, K, or halogens were removed from the set, as
were molecules containing chains of more than five sequential
carbon atoms, azides, aldehydes, and hydrazones, and molecules
with MW > 350. For each reagent, the physiologically relevant
protonation states were selected.
The de novo drug design program Skelgen was used for the
virtual screen. A brief discussion of the Skelgen algorithm is given
below, and further details of the principles involved as well as
practical applications of the method can be found in previous
publications.37,43-45 Skelgen takes as input the three-dimensional
coordinates of the protein, a rectangular box that defines the binding
site pocket, and a predefined set of fragments. The algorithm
invokes a simulated annealing procedure to minimize a function
that assesses the suitability of molecular structures within the active
site box. In this process, a set of initial fragments are randomly
selected and connected together. The putative ligand is then altered
by several transitions, which include rigid-body displacements,
conformational changes, and fragment addition, deletion, and
replacement. The transition type is randomly chosen, and at each
step the value of the objective function is determined. The modified
structure is accepted or rejected on the basis of the Metropolis
condition, p ) exp(∆F/T), where ∆F represents the change in value
of the objective function resulting from the transition, T is the
temperature, which is lowered during the annealing procedure, and
p is the probability of acceptance of the transition. The length of
each Skelgen annealing run is defined by the number of transitions
at each temperature setting, termed the Markov chain length, and
by the number of Markov chains as the temperature is depressed.
The algorithm generates one structure for each run, and as a result
of the stochastic nature of the simulation, different random starts
can lead to different solutions.
General Procedure C: 3-Alkoxy-2,3-dihydroisoindol-1-ones
(NIS Method). The appropriate 3-benzylsulfanylisoindol-1-one (1
mol equiv) in THF (4 mL) was added to a solution of NIS (1.1
mol equiv), CSA (0.1 mol equiv), and the appropriate alcohol (2.2
mol equiv) in THF (3 mL). The reaction was stirred in the dark for
4 h at room temperature and then concentrated in vacuo. The residue
was dissolved in EtOAc (30 mL) and washed with aqueous sodium
thiosulfate (2 × 30 mL). The organic layer was collected, dried
(Na2SO4), and concentrated in vacuo to give the 3-alkoxy-isoindol-
1-one.
2-Benzyl-3-(4-chlorophenyl)-3-(3-hydroxypropoxy)-2,3-dihy-
droisoindol-1-one (76). General Procedure C. 12b (0.50 g, 1.10
mmol) in THF (4 mL), NIS (0.27 g, 1.21 mmol), CSA (0.03 g),
and propane-1,3-diol (0.17 mL, 2.35 mmol) in THF (3 mL) was
purified by chromatography (silica gel, 40:60 EtOAc/petrol),
followed by recrystallization (EtOAc, petroleum ether) to give 76
as a white solid (0.17 g, 0.42 mmol, 39%); mp 149-151 °C. IR
(Diamond ATR) υmax (cm-1): 3431 (OH), 1669 (CO), 1426, 1399,
1
1359, 1066, 1012, 818, 766, 700. H NMR (300 MHz, CDCl3) δ
1.21-1.43 (2H, m, OCH2CH2CH2OH), 1.53 (s, CH2OH, ex), 2.69-
2.74 (2H, m, CH2OH), 3.40-3.44 (2H, m, OCH2CH2CH2O), 3.89-
3.94 (1H, d, J ) 14.5 Hz, NCH2), 4.69-4.74 (1H, d, J ) 14.5 Hz,
NCH2), 7.01-7.10 (1H, m, Ar-H), 7.12-7.23 (9H, m, Ar-H), 7.40-
7.46 (2H, m, Ar-H), 7.82-7.88 (1H, m, isoindolinone-H). 13C NMR
(300 MHz, CDCl3) δ 32.02 (OCH2CH2CH2OH), 43.35 (NCH2),
60.70 and 60.83 (OCH2CH2CH2OH), 95.38 (C-3), 123.11, 124.04,
127.58, 128.16, 128.53, 128.96, 129.55, 130.15, 131.81, 133.08,
134.73, 137.50, 137.89, 145.51, 165.76 (C-1). LC-MS (ESI+) m/z
332, 334, 408 [M + H]+. Anal. (C24H22ClNO3) C, H, N.
In this work, Skelgen was used in reagent screening mode, such
that each reagent in our data set was attached to a fixed isoindoli-
none core and individually subject to a Skelgen simulated annealing
run, and thus, transition types for fragment addition, deletion, and
replacement were switched off, as were the translation and rotation
of the whole ligand. Only transition types for conformational
changes of the attached reagent were retained. The objective
function comprises several terms, which are conferred a zero value
General Procedure D: Synthesis of Isoindolin-1-ones Deriva-
tives with R4 Amino Substitutions. A solution of the appropriate
3-hydroxyisoindolinone 10 (1.0 equiv) in THF (10 mL) was treated
with a solution of thionyl chloride (2.2 equiv) and a catalytic amount
of DMF. The mixture was stirred 16 h and then concentrated in

Isoindolinone MDM2-p53 Inhibitors
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 21 6219
vacuo. The residues were dissolved in either DMF (10 mL) or
acetonitrile (10 mL), as appropriate, and treated with the appropriate
primary amine (1.1 equiv) and either anhydrous potassium carbonate
(2.2 equiv) or triethylamine (2.2 equiv). The mixture was stirred
at room temperature under a nitrogen atmosphere, (EtOAc/petrol,
3:2). After 20 h, the solvent was removed in vacuo. The crude
product was extracted with EtOAc (15 mL) and water (20 mL).
The organic layers were combined, dried (Na2SO4), and concen-
trated in vacuo. Purification was done by flash chromatography
(EtOAc/petrol, 1:4) and by recrystallization from suitable solvents.
General Procedure E: Synthesis of Isoindolin-1-ones Deriva-
tives with R4 Alkoxy Substitution. A solution of the appropriate
3-hydroxyisoindolinone 10 (1.0 equiv) in THF (10 mL) was treated
with a solution of thionyl chloride (2.2 equiv) and a catalytic amount
of DMF. After 16 h, the mixture was concentrated in vacuo. The
residues were dissolved in either DMF (5-10 mL) or THF (5-10
mL), as appropriate, and treated with the appropriate primary
alcohol (1.1 equiv or 2.2 equiv) with or without triethylamine (2.2
equiv). The reaction mixture was stirred at room temperature for
20 h, and the workup was performed as for general procedure D.
3-(4-Chlorophenyl)-3-(4-hydroxy-3,5-dimethoxybenzyloxy)-
2-propyl-2,3-dihydroisoindol-1-one (79). General Procedure E:
10k (250 mg, 0.82 mmol) and syringic alcohol (331 mg, 1.80 mmol)
were purified by chromatography (45% EtOAc, petroleum ether)
and HPLC (Genesis C18; H2O/CH3CN, 270 nm) to give 79 as a
1/200 dilution), then washed three times with TBS-Tween before
incubation for 45 min at 20 °C with a goat-anti-mouse horseradish
peroxidase (HRP) conjugated secondary antibody (Dako, used at
1/2000). The unbound secondary antibody was removed by washing
three times with TBS-Tween. The bound HRP activity was
measured by enhanced chemiluminescence (ECL, Amersham
Biosciences) using the oxidation of the diacylhydrazide substrate,
luminol, to generate a quantifiable light signal. The luminol substrate
together with enhancer was automatically injected into each well
and the relative luminescence units (RLU) measured over a 30 s
interval using a Berthold MicroLumat-Plus LB 96 V microplate
luminometer. The percentage MDM2 inhibition at a given concen-
tration is calculated as the (RLU detected in the compound treated
sample ÷ RLU of DMSO controls) × 100. The IC50 was calculated
using a plot of % MDM2 inhibition versus concentration and is
the average of three independent experiments. The results (mean
( S.D.) are presented in Table 1.
Cell-Based Assays. Western Blot Analysis. SJSA cells were
treated for 6 h with 5, 10, and 20 µM of compounds 76 and 79 in
1% DMSO. The cells together with 1% DMSO only controls were
washed with ice-cold phosphate buffered saline (PBS) and protein
extracts prepared by lysing the cells in SDS buffer (62.5 mM Tris
pH 6.8; 2% sodium dodecyl sulfate (SDS); 10% glycerol) with
sonication for 2 × 5 s (Soniprep 150ME) to break down high
molecular weight DNA and reduce the viscosity of the samples.
The protein concentration of the samples was estimated using the
Pierce BCA assay system (Pierce, Rockford, IL) and 50 µg aliquots
of protein analyzed using standard SDS-polyacrylamide gel elec-
trophoresis (SDS-PAGE) and western immunoblotting procedures.
â-Mercaptoethanol (5%) and bromophenol blue (0.05%) were added
and the samples, which were then boiled for 5 min, followed by
brief centrifugation, before loading onto a precast 4-20% gradient
Tris-glycine buffered SDS-polyacrylamide gel (Invitrogen). Mo-
lecular weight standards (SeeBlue, Invitrogen) were included on
every gel, and electrophoresis was carried out in a Novex XL tank
(Invitrogen) at 180 V for 90 min. The separated proteins were
transferred electrophoretically overnight from the gel onto a Hybond
C nitrocellulose membrane (Amersham) using a BioRad electro-
phoresis tank and 25 mM Tris, 190 mM glycine, and 20% methanol
transfer buffer at 30 V. Primary antibodies used for immuno-
detection of the transferred proteins were: mouse monoclonal NCL-
p53DO-7 (Novocastra) at 1:1000; MDM2 (Ab-1, clone IF2;
Oncogene) at 1:500; WAF1 (Ab-1, clone 4D10; Oncogene) at
1:100; Actin (AC40; Sigma) at 1:1000. The secondary antibody
used was peroxidase conjugated, affinity purified, goat anti-mouse
(Dako) at 1:1000. Protein detection and visualization was performed
by ECL (Amersham) with light detection by exposure to blue-
sensitive autoradiography film (Super RX, Fuji).
light red oil (180 mg, 0.38 mmol, 46%). λ
(CH3OH)/nm 209,
max
Abs 0.550. IR: 3360, 2933, 1692, 1604, 1504, 1450 cm-1. 1H NMR
(300 MHz, CDCl3) δ 0.74 (t, 3H, J ) 7.4 Hz, CH2CH2CH3), 1.31
(m, 1H, NCH2CH2), 1.44 (m, 1H, NCH2CH2), 3.03 (m, 1H, NCH2),
3.23, (m, 1H, NCH2), 3.79 (s, 6H, OMe), 3.84 (d, 1H, J ) 11.1
Hz, OCH2), 4.08 (d, 1H, J ) 11.2 Hz, OCH2), 5.45 (s, 1H, OH),
6.38 (s, 2H, Ar-H), 7.05 (m, 1H, Ar-H), 7.22 (d, 2H, J ) 8.9 Hz,
Ar-H), 7.28 (d, 2H, J ) 8.7 Hz, Ar-H), 7.42 (m, 2H, Ar-H), 7.83
(m, 1H, Ar-H). 13C NMR (75 MHz, CDCl3) δ 12.1, 22.1, 41.9,
56.7, 65.6, 95.1, 104.8, 123.5, 123.8, 128.2, 128.7, 129.1, 130.2,
132.3, 132.8, 134.7, 134.8, 138, 145.5, 147.3, 168.6. LC/MS (ESI+)
m/z 302.1, 489.9, 500. Anal. (C26H26ClNO5‚0.1C4H8O2) C, H, N.
MDM2-p53 interaction using a 96-well plate binding assay
(ELISA). The 96-well black and white high binding luminometry
isoplates (Wallac, Cat N0 140-155) were coated by overnight
incubation at 35 °C with 200 µL per well of 5 µg mL-1 streptavidin
(Chemicon International) in coating buffer (0.1 M Na2HPO4‚2H2O;
0.1 M citric acid; pH 5.0). The plates were washed five times in
1× dissociation enhanced lanthanide fluorescence immunoassay
(DELFIA) buffer (Wallac) and then incubated for 3 h at room
temperature with saturation buffer (0.3 M D-sorbitol; 50 mM Tris;
150 mM NaCl; 0.1% BSA; 0.05% sodium azide; pH 7.0) to block
nonspecific protein binding sites on the plate. After removal of the
buffer from the plates, they were allowed to dry in a sterile laminar
air flow hood at room temperature before incubation for 1 h at 4
°C with 200 µL per well of 100 µg mL-1 biotinylated IP3 peptide
(b-IP3: Ac-Met-Pro-Arg-Phe19-Met-Asp-Tyr-Trp-Glu-Gly-Leu26-
Asn-NH2)17 dissolved in 0.05% DMSO-PBS, pH 7.4 buffer. After
washing the wells three times with PBS, the plates were ready to
use for MDM2 binding.
Results of the analysis by western immunoblotting for p53,
p21WAF1, and actin are shown in Figure 1. A dose-dependent
increase in MDM2 and p21WAF1 was observed in MDM2 amplified
SJSA cells treated with compounds 76 and 79, consistent with the
release of transcriptionally active p53. No change was observed
for p53 protein levels, although there was clear evidence of
increased MDM2 and p21WAF1 expression.
For initial testing, the compounds and controls were plated out
in triplicate into clear 96-well plates (Nunc) in 10-µL aliquots to
give final concentrations of 500 µM, 100 µM, and 20 µM in the
assay. Control samples consisted of 5% DMSO carrier alone as a
negative control and 100 nM active peptide (AP-B: Ac-Phe19-Met-
Aib-Pmp-6-Cl-Trp-Glu-Ac3-Leu26-NH2) as a positive control pep-
tide antagonist of the MDM2-p53 interaction (IC50 ) 5 nM).18
Compounds and controls aliquoted in 96-well plates were pre-
incubated at 20 °C for 20 min with 190 µL aliquots of optimized
concentrations of in vitro translated MDM2, before transfer of the
MDM2-compound mixture to the b-IP3 streptavidin plates, and
incubation at 4 °C for 90 min. After washing three times with PBS
to remove unbound MDM2, each well was incubated at 20 °C for
1 h with a TBS-Tween (50 mM Tris pH 7.5; 150 mM NaCl; 0.05%
Tween 20 nonionic detergent) buffered solution of primary mouse
monoclonal anti-MDM2 antibody (Ab-5, Calbiochem, used at a
p53-Dependent Reporter Gene Assay. The reporter gene assay
used in this investigation involved the transient transfection of a
plasmid luciferase reporter gene construct, pLubP2, which was made
from a pGL3 plasmid (Invitrogen) into which a luciferase gene
had been introduced downstream of the p53 responsive MDM2 P2
promoter.49 The introduction of this plasmid into cells results in
the expression of luciferase in a p53-dependent manner and, hence,
an increase in the transcriptional activity of p53 is reflected in
increased levels of luciferase. SJSA cells were seeded into 96-well
clear bottom luminometer plates (Wallac) in RPMI medium
(GIBCO) supplemented with 5% fetal bovine serum and allowed
to attach overnight. Plasmid transfection was carried out using the
FuGENE 6 transfection reagent (Roche Molecular Biochemicals).
Each 25 µL transfection volume per well included 30 ng of the
pLubP2 reporter plasmid or pGL3-basic empty vector as a negative
control and 5 ng of pCMVâ-galactosidase (Stratagene) as a

6220 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 21
Hardcastle et al.
(15) Becattini, B.; Sareth, S.; Zhai, D. Y.; Crowell, K. J.; Leone, M.;
Reed, J. C.; Pellecchia, M. Targeting apoptosis via chemical design:
Inhibition of bid-induced cell death by small organic molecules.
Chem. Biol. 2004, 11, 1107-1117.
(16) Kussie, P. H.; Gorina, S.; Marechal, V.; Elenbaas, B.; Moreau, J.;
Levine, A. J.; Pavletich, N. P. Structure of the MDM2 oncoprotein
bound to the p53 tumor suppressor transactivation domain. Science
1996, 274, 948-953.
(17) Bottger, V.; Bottger, A.; Howard, S. F.; Picksley, S. M.; Chene, P.;
Garcia-Echeverria, C.; Hochkeppel, H. K.; Lane, D. P. Identification
of novel mdm2 binding peptides by phage display. Oncogene 1996,
13, 2141-2147.
(18) Garcia-Echeverria, C.; Chene, P.; Blommers, M. J. J.; Furet, P.
Discovery of potent antagonists of the interaction between human
double minute 2 and tumor suppressor p53. J. Med. Chem. 2000,
43, 3205-3208.
(19) Stoll, R.; Renner, C.; Hansen, S.; Palme, S.; Klein, C.; Belling, A.;
Zeslawski, W.; Kamionka, M.; Rehm, T.; Muhlhahn, P.; Schumacher,
R.; Hesse, F.; Kaluza, B.; Voelter, W.; Enhh, R. A.; Holak, T. A.
Chalcone derivatives antagonise interactions between the human
oncoprotein MDM2 and p53. Biochemistry 2001, 40, 336-344.
(20) Zhao, J. H.; Wang, M. J.; Chen, J.; Luo, A. P.; Wang, X. Q.; Wu,
M.; Yin, D. L.; Liu, Z. H. The initial evaluation of nonpeptidic small-
molecule HDM2 inhibitors based on p53-HDM2 complex structure.
Cancer Lett. 2002, 183, 69-77.
(21) Majeux, N.; Scarsi, M.; Caflisch, A. Efficient electrostatic solvation
model for protein-fragment docking. Proteins 2001, 42, 256-
268.
(22) Duncan, S. J.; Gruschow, S.; Williams, D. H.; McNicolas, C.;
Purewal, R.; Hajek, M.; Gerlitz, M.; Martin, S.; Wrigley, S. K.;
Moore, M. Isolation and structure elucidation of chlorofusin, a novel
p53-MDM2 antagonist from a Fusarium sp. J. Am. Chem. Soc. 2001,
123, 554-560.
(23) Chene, P.; Fuchs, J.; Bohn, J.; Garcia-Echeverria, C.; Furet, P.;
Fabbro, D. A small synthetic peptide, which inhibits the p53-hdm2
interaction, stimulates the p53 pathway in tumour cell lines. J. Mol.
Biol. 2000, 299, 245-253.
(24) Vassilev, L. T.; Vu, B. T.; Graves, B.; Carvajal, D.; Podlaski, F.;
Filipovic, Z.; Kong, N.; Kammlott, U.; Lukacs, C.; Klein, C.;
Fotouhi, N.; Liu, E. A. In vivo activation of the p53 pathway
by small-molecule antagonists of MDM2. Science 2004, 303, 844-
848.
(25) Grasberger, B. L.; Lu, T. B.; Schubert, C.; Parks, D. J.; Carver, T.
E.; Koblish, H. K.; Cummings, M. D.; LaFrance, L. V.; Milkiewicz,
K. L.; Calvo, R. R.; Maguire, D.; Lattanze, J.; Franks, C. F.; Zhao,
S. Y.; Ramachandren, K.; Bylebyl, G. R.; Zhang, M.; Manthey, C.
L.; Petrella, E. C.; Pantoliano, M. W.; Deckman, I. C.; Spurlino, J.
C.; Maroney, A. C.; Tomczuk, B. E.; Molloy, C. J.; Bone, R. F.
Discovery and cocrystal structure of benzodiazepinedione HDM2
antagonists that activate p53 in cells. J. Med. Chem. 2005, 48, 909-
912.
(26) Parks, D. J.; LaFrance, L. V.; Calvo, R. R.; Milkiewicz, K. L.; Gupta,
V.; Lattanze, J.; Ramachandren, K.; Carver, T. E.; Petrella, E. C.;
Cummings, M. D.; Maguire, D.; Grasberger, B. L.; Lu, T. B. 1,4-
Benzodiazepine-2,5-diones as small molecule antagonists of the
HDM2-p53 interaction: discovery and SAR. Bioorg. Med. Chem.
Lett. 2005, 15, 765-770.
(27) Geiger, T.; Husken, D.; Weiler, J.; Natt, F.; Woods-Cook, K. A.;
Hall, J.; Fabbro, D. Consequences of the inhibition of Hdm2
expression in human osteosarcoma cells using antisense oligo-
nucleotides. Anti-Cancer Drug Des. 2000, 15, 423-430.
(28) Meye, A.; Wurl, P.; Bache, M.; Bartel, F.; Grunbaum, U.; Mansa-
ard, J.; Schmidt, H.; Taubert, H. Colony formation of soft tissue
sarcoma cells is inhibited by lipid-mediated antisense oligodeoxy-
nucleotides targeting the human mdm2 oncogene. Cancer Lett. 2000,
149, 181-188.
(29) Chen, L.; Agrawal, S.; Zhou, W.; Zhang, R.; Chen, J. Synergistic
activation of p53 by inhibition of MDM2 expression and DNA
damage. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 195-200.
(30) Wasylyk, C.; Salvi, R.; Argentini, M.; Dureuil, C.; Delumeau, I.;
Abecassis, J.; Debussche, L.; Wasylyk, B. p53 mediated death of
cells overexpressing MDM2 by an inhibitor of MDM2 interaction
with p53. Oncogene 1999, 18, 1921-1934.
(31) Chene, P.; Fuchs, J.; Carena, I.; Furet, P.; Echeverria, C. G. Study
of the cytotoxic effect of a peptidic inhibitor of the p53-hdm2
interaction in tumor cells. FEBS Lett. 2002, 529, 293-297.
(32) Hardcastle, I. R.; Ahmed, S. U.; Atkins, H.; Farnie, G.; Golding, B.
T.; Griffin, R. J.; Guyenne, S.; Hutton, C.; Ka¨llblad, P.; Kemp, S.
J.; Kitching, M. S.; Newell, D. R.; Norbedo, S.; Northen, J. S.; Reid,
R. J.; Saravanan, K.; Willems, H. M. G.; Lunec, J. Isoindolinone
based inhibitors of the MDM2-p53 protein-protein interaction.
Bioorg. Med. Chem. Lett. 2005, 15, 1515-1520.
transfection efficiency control per well, together with 0.2 µL of
FuGENE 6, and the volume made up to 25 µL with serum-free
medium.
Following 24 h transfection, the cells were treated with the
MDM2 inhibitors, which were made up in cell culture medium to
the appropriate concentration from concentrated stocks dissolved
in DMSO. Five replicate wells were used for each inhibitor
concentration. In addition, DMSO controls and untreated wells were
included. Following the incubation period with the inhibitors, the
96-well plates were checked under a light microscope for possible
lifting of cells. All the experiments carried out resulted in no
detectable cell lifting. The medium from the wells was gently
aspirated, and 25 µL of lysis buffer (Applied Biosystems) was added
into each well. The plate was then shaken on a rocker for 5 min
and stored at -20 °C until the luminometry was carried out.
Luciferase and â-galactosidase activities were measured using a
dual light chemiluminescence detection kit (Applied Biosystems)
and a luminometer (LB 96 V, EG&G Berthold) to quantify the
light emission. Normalized luciferase activity levels are calculated
as a ratio of luciferase/â-galactosidase and fold-induction with the
compounds calculated relative to the background with DMSO
solvent control alone.
Acknowledgment. The authors thank Cancer Research UK,
BBSRC, and EPSRC for funding. The EPSRC Mass Spectrom-
etry Service at the University of Wales (Swansea) is gratefully
acknowledged.
Supporting Information Available: Additional details of the
first round of virtual screening, experimental details and analytical
data for all compounds and intermediates, and a table of combustion
analysis data. This material is available free of charge via the
Internet at http://pubs.acs.org.
References
(1) Lane, D. P. Cancer. p53, guardian of the genome. Nature 1992, 358,
15-16.
(2) Vousden, K. H.; Lu, X. Live or let die: the cell’s response to p53.
Nat. ReV. Cancer 2002, 2, 594-604.
(3) Momand, J.; Zambetti, G. P.; Olson, D. C.; George, D.; Levine, A.
The mdm-2 oncogene product forms a complex with p53 protein and
inhibits p53-mediated transactivation. Cell 1992, 69, 1237-1245.
(4) Fuchs, S. Y.; Adler, V.; Buschmann, T.; Wu, X. W.; Ronai, Z. Mdm2
association with p53 targets its ubiquitination. Oncogene 1998, 17,
2543-2547.
(5) Oliner, J. D.; Kinzler, K. W.; Meltzer, P. S.; George, D. L.;
Vogelstein, B. Amplification of a gene encoding a p53-associated
protein in human sarcomas. Nature 1992, 358, 80-83.
(6) Chene, P. Inhibiting the p53-MDM2 interaction: An important target
for cancer therapy. Nat. ReV. Cancer 2003, 3, 102-109.
(7) Cochran, A. G. Antagonists of protein-protein interactions. Chem.
Biol. 2000, 7, R85-R94.
(8) Cochran, A. G. Protein-protein interfaces: mimics and inhibitors.
Curr. Opin. Chem. Biol. 2001, 5, 654-659.
(9) Toogood, P. L. Inhibition of protein-protein association by small
molecules: Approaches and progress. J. Med. Chem. 2002, 45, 1543-
1558.
(10) Yin, H.; Hamilton, A. D. Strategies for targeting protein-protein
interactions with synthetic agents. Angew. Chem., Int. Ed. 2005, 44,
4130-4163.
(11) Braisted, A. C.; Oslob, J. D.; Delano, W. L.; Hyde, J.; McDowell,
R. S.; Waal, N.; Yu, C.; Arkin, M. R.; Raimundo, B. C. Discovery
of a potent small molecule IL-2 inhibitor through fragment assembly.
J. Am. Chem. Soc. 2003, 125, 3714-3715.
(12) Yin, H.; Lee, G. I.; Sedey, K. A.; Kutzki, O.; Park, H. S.; Omer, B.
P.; Ernst, J. T.; Wang, H. G.; Sebti, S. M.; Hamilton, A. D. Terphenyl-
based bak BH3 alpha-helical proteomimetics as low-molecular-weight
antagonists of Bcl-X-L. J. Am. Chem. Soc. 2005, 127, 10191-10196.
(13) Lesuisse, D.; Lange, G.; Deprez, P.; Benard, D.; Schoot, B.; Delettre,
G.; Marquette, J. P.; Broto, P.; Jean-Baptiste, V.; Bichet, P.; Sarubbi,
E.; Mandine, E. SAR and X-ray. A new approach combining
fragment-based screening and rational drug design: Application to
the discovery of nanomolar inhibitors of Src SH2. J. Med. Chem.
2002, 45, 2379-2387.
(14) Zeng, L.; Li, J. M.; Muller, M.; Yan, S.; Mujtaba, S.; Pan, C. F.;
Wang, Z. Y.; Zhou, M. M. Selective small molecules blocking HIV-1
Tat and coactivator PCAF association. J. Am. Chem. Soc. 2005, 127,
2376-2377.

Isoindolinone MDM2-p53 Inhibitors
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 21 6221
(33) Paull, K. D.; Shoemaker, R. H.; Hodes, L.; Monks, A.; Scudiero, D.
A.; Rubinstein, L.; Plowman, J.; Boyd, M. R. Display and analysis
of patterns of differential activity of drugs against human tumor cell
lines: development of mean graph and COMPARE algorithm. J. Natl.
Cancer Inst. 1989, 81, 1088-1092.
(41) Orndorff, W. R.; Murray R. R. A new class of phthaleinssmixed
phthaleinssformed by heating p-hydroxy-o-benzoic acids with
phenols. J. Am. Chem. Soc. 1917, 39, 679-697.
(42) Maple, J. R.; Dinur, U.; Hagler, A. T. Derivation of force-fields for
molecular mechanics and dynamics from ab initio energy surfaces.
Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 5350-5354.
(34) Mancera, R. L.; Ka¨llblad, P.; Todorov, N. P. Ligand-protein docking
using a quantum stochastic tunneling optimization method. J. Comput.
Chem. 2004, 25, 858-864.
(43) Todorov, N. P.; Dean, P. M. Evaluation of a method for controlling
molecular scaffold diversity in de novo ligand design. J. Comput.-
Aided Mol. Des. 1997, 11, 175-192.
(44) Kallblad, P.; Todorov, N. P.; Willems, H. M. G.; Alberts, I. L.
Receptor flexibility in the in silico screening of reagents in the S1′
pocket of human collagenase. J. Med. Chem. 2004, 47, 2761-2767.
(45) Firth-Clark, S.; Willems, H. M. G.; Williams, A.; Harris, W.
Generation and selection of novel estrogen receptor ligands using
the de novo structure-based design tool, SkelGen. J. Chem. Inf.
Comput. Sci. 2006, 46, 642-647.
(46) Bondi, A. Van der Waals volumes and radii. J. Phys. Chem. 1968,
68, 441-451.
(47) Bohm, H. J. The development of a simple empirical scoring function
to estimate the binding constant for a protein ligand complex of
known 3-dimensional structure. J. Comput.-Aided Mol. Des. 1994,
8, 243-256.
(48) Todorov, N. P.; Dean, P. M. A branch-and-bound method for optimal
atom-type assignment in de novo ligand design. J. Comput.-Aided
Mol. Des. 1998, 12, 335-349.
(49) Liang, H.; Lunec, J. Characterisation of a novel p53 down-regulated
promoter intron 3 of the human MDM2 oncogene. Gene 2005, 361,
112-118.
(35) Ka¨llblad, P.; Mancera, R. L.; Todorov, N. P. Assessment of multiple
binding modes in ligand-protein docking. J. Med. Chem. 2004, 47,
3334-3337.
(36) Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R.
Development and validation of a genetic algorithm for flexible
docking. J. Mol. Biol. 1997, 267, 727-748.
(37) Stahl, M.; Todorov, N. P.; James, T.; Mauser, H.; Boehm, H. J.;
Dean, P. M. A validation study on the practical use of automated de
novo design. J. Comput.-Aided Mol. Des. 2002, 16, 459-478.
(38) Kitching, M. S.; Clegg, W.; Elsegood, M. R. J.; Griffin, R. J.;
Golding, B. T. Synthesis of 3-alkoxy and 3-alkylamino-2-alkyl-3-
arylisoindolinones. Synlett 1999, 997-999.
(39) Sayle, K. L.; Bentley, J.; Boyle, F. T.; Calvert, A. H.; Cheng, Y. Z.;
Curtin, N. J.; Endicott, J. A.; Golding, B. T.; Hardcastle, I. R.;
Jewsbury, P.; Mesguiche, V.; Newell, D. R.; Noble, M. E. M.;
Parsons, R. J.; Pratt, D. J.; Wang, L. Z.; Griffin, R. J. Structure-
based design of 2-arylamino-4-cyclohexylmethyl-5-nitroso-6-
aminopyrimidine inhibitors of cyclin-dependent kinases 1 and 2.
Bioorg. Med. Chem. Lett. 2003, 13, 3079-3082.
(40) Douglas, N. L.; Ley, S. V.; Lucking, U.; Warriner, S. L. Tuning
glycoside reactivity: New tool for efficient oligosaccharide synthesis.
J. Chem. Soc., Perkin Trans. 1 1998, 51-65.
JM0601194