Enantiomerically pure 1,4-benzodiazepine-2,5-diones as Hdm2 antagonists

Article abstract of DOI:10.1016/j.bmcl.2006.03.067

The 1,4-benzodiazepine-2,5-dione is a suitable template to disrupt the interaction between p53 and Hdm2. The development of an enantioselective synthesis disclosed the stereochemistry of the active enantiomer. An in vitro p53 peptide displacement assay identified active compounds. These activities were confirmed in several cell-based assays including induction of the p53 regulated gene (PIG-3) and caspase activity.

Full text of DOI:10.1016/j.bmcl.2006.03.067

Bioorganic & Medicinal Chemistry Letters 16 (2006) 3115–3120
Enantiomerically pure 1,4-benzodiazepine-2,5-diones as
Hdm2 antagonists
Juan Jose Marugan,* Kristi Leonard, Pierre Raboisson, Joan M. Gushue, Raul Calvo,
Holly K. Koblish, Jennifer Lattanze, Shuyuan Zhao, Maxwell D. Cummings,
Mark R. Player, Carsten Schubert, Anna C. Maroney and Tianbao Lu
Drug Discovery, Johnson & Johnson Pharmaceutical Research and Development, L.L.C., 665 Stockton Drive,
Exton, PA 19341, USA
Received 14 February 2006; revised 20 March 2006; accepted 20 March 2006
Available online 21 April 2006
Abstract—The 1,4-benzodiazepine-2,5-dione is a suitable template to disrupt the interaction between p53 and Hdm2. The develop-
ment of an enantioselective synthesis disclosed the stereochemistry of the active enantiomer. An in vitro p53 peptide displacement
assay identified active compounds. These activities were confirmed in several cell-based assays including induction of the p53
regulated gene (PIG-3) and caspase activity.
Ó 2006 Elsevier Ltd. All rights reserved.
P53 is a nuclear transcription factor that acts as a tumor
suppressor. P53 mutations are very common in human
tumors, however it remains wild-type in approximately
50% of human cancers.1 DNA damage due to chemo-
therapy or radiation increases the levels of p53, promot-
ing the synthesis of proteins involved in DNA repair.2
As a consequence, elevation of p53 levels assists in cell
cycle arrest and/or apoptosis. However, high levels of
p53 also leads to increased expression of the human
homolog of the double minute-2 protein (Hdm2).3a
Hdm2 is an E3 ubiquitin ligase and when bound to
p53 leads to p53 ubiquination and degradation by the
proteasome complex, thus facilitating resistance to che-
motherapy or radiation attack.3b It is also known that
Hdm2 over-expression is associated with poor clinical
outcome for some cancers.4 Therefore, disruption of
the interaction between these two proteins is an appeal-
ing target for chemotherapeutic treatment of wild-type
p53 tumors.
1,4-benzodiazepine-2,5-dione as a suitable template for
inhibiting the Hdm2:p53 interaction.
Previously we described the synthesis and SAR of novel
racemic 1,4-benzodiazepinediones with improved cellu-
lar activity.6e Here, we describe the synthesis of enantio-
merically pure 1,4-benzodiazepinediones as Hdm2
antagonists, as well as some of the biological activities
obtained with these compounds.
In order to evaluate the stereochemical effects on the
activity of our inhibitors, we developed a new methodol-
ogy for the synthesis of enantiomerically pure
compounds.
The synthetic strategy, outlined in Scheme 1 and 2,
consisted of the preparation and selective crystalliza-
tion of the diastereomeric camphanic esters 3 and 4.
Compound 1 was treated with methyl lithium to give
the corresponding racemic alcohol 2. This was treated
with camphanic chloride, followed by selective recrys-
tallization to give 4. After separation and hydrolysis,
the enantiomerically pure alcohol was transformed to
the corresponding amine 7 via Mitsunobu reaction
with phthalimide and deprotection with hydrazine. In
2, amine 7 was subjected to an Ugi reaction followed
by in situ cyclization to generate a 1:1 mixture of the
1,4-benzodiazepine-2,5-dione templates 10 and 11,
Several authors have described Hdm2 antagonists.5 Our
group,6 has published a number of papers describing the
Keywords: Hdm2; Protein–protein interaction; p53.
*
Corresponding author. Tel.: +1 484 888 8257; e-mail addresses:
jmaruga@prdus.jnj.com; Paoyjuan@aol.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.03.067

3116
J. J. Marugan et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3115–3120
Cl
Cl
Cl
O
NO2
NO2
O
OH
O
+
i
O
O
H
O
O
O
Cl
NO2
Cl
NO2
ii
O
O
O
O
1
2
3
4
Cl
Cl
Cl
Cl
NO2
NO2
iii
O
v
iv
O
N
NO2
NO2
O
O
HO
H2N
O
O
4
5
6
7
Scheme 1. Reagents and conditions: (i) MeLi, THF, À78 °C to rt, 30 min, 73%; (ii) (1S)-(-)-camphanic chloride, NEt3, CH2Cl2, 16 h, 98%; (iii) LiOH,
MeOH–H2O, 16 h, 93%; (iv) phthalimide, Ph3P, DIAD, THF, rt, 3 h, 87%; (v) hydrazine, THF, 80 °C, 1 h, 96%.
O
The stereochemistry of 12 was assigned on the basis of
small molecule X-ray crystallography, conformational
analysis and comparison with the structure of a closely
related analog bound to Hdm2.6a
Cl
I
OH
NH
O
vi
H
vii
NO2
O
O
Cl
Small molecule crystallography of compound 4 (Fig. 1),
gave an unambiguous assignment of the stereochemistry
at the substituted chiral nitrogen.7 Conformational anal-
ysis with the two possible diastereomers at position 3
established a strong preference for one low energy con-
former in each case (not shown) Figure 2. Comparison
of the preferred conformer for each of the two diastereo-
mers with the published crystal structure of an HDM2-
bound analog in this series (Ref. 6a) established that one
conformer was essentially identical to that of the close
analog bound to HDM2 (Figs. 3 and 4 and additional
discussion below) while the other could not fulfill the
same binding interactions. On this basis the stereochem-
istry at position 3 was assigned.
H2N
7
8
9
Cl
Cl
NO2
NO2
O
O
+
I
I
N
N
Cl
Cl
N
H
N
H
O
O
10
11
Cl
Cl
NO2
NH2
viii
ix
O
O
I
I
N
N
Cl
Cl
N
H
N
R1
O
O
10
12
Scheme 2. Reagents and conditions: (vi) MeOH, 1-isocyanocyclohex-
ene, rt, 2 days; (vii) AcCl, rt to 60 °C, 1 h.,73%; (viii) X-R1, Bu4NI,
K2CO3, THF, 80 °C, 2 h, 40–95%; (ix) NH4Cl, Zn(0), AcOH.
which were readily separated by column chromatogra-
phy. Alkylation of diastereoisomer 10 followed by
reduction of the nitro group yielded the target Hdm2
antagonists 12.
Figure 1. X-ray structure determination of compound 4.

J. J. Marugan et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3115–3120
3117
As mentioned in our previous papers,6b,e the introduc-
tion of an amine functional group in the ortho position
of the benzylic ring significantly improved the IC50
values, but dramatically reduced the cellular activity.
The reduction of cellular potency is probably due to
the zwitterionic character of the molecule. A binding
model of an Hdm2-antagonist complex was used to
guide the exploration of solubilizing modifications of
this series of molecules.
O
2
R
4
R
N
3
R
N
1
O
R
Figure 2.
C
Gln72
O
I
Based on the previously disclosed structure of a closely
related antagonist bound to Hdm2,6a we developed a
binding model for the complex schematized in Figure 3.
The binding interaction involves three lipophilic pock-
ets that are complemented by the three aromatic rings
of the bound benzodiazepine. The binding model pre-
dicts that the ortho-aniline is well positioned to hydro-
gen bond with the backbone carbonyl of Val93 (Figs.
3 and 4). This prediction was validated by the ob-
served activity of various ortho-anilines in this series
(12e, 12i, and 12k).
solvent exposed
O
HN
O
N
C
NH2
O
Val93
C
O
Leu54
Cl
Cl
Figure 3. Binding model schematic for the complex of 1,4-benzodiaz-
epine-2,5-diones bound to Hdm2.
The binding model and related crystal structures (not
shown) also indicate that the aliphatic acid solubilizing
group is relatively solvent-exposed (Fig. 4). We did not
expect a strong specificity based on intermolecular bind-
ing interactions with substituents at this position.
Analysis of SAR reported in Table 1 shows that the
replacement of the valeric acid with the morpholin-4-
ylethyl or with 2-(2-methoxyethoxy) ethyl solubilizing
chains negatively impacts both the affinity measured in
the FP1 binding assay and cellular PIG3 production
(compare 12b with 12c and 12g). Different results were
observed with the corresponding amino analogues
(12a, 12e and 12i), which exhibit approximately the
same affinity in the FP1 binding assay. Moreover, the
non-acidic compounds 12e and 12i proved to perform
better in the PIG3 cell-based assay than the valeroyl
derivative 12a. Although the PIG3 production was also
increased when the 1-methyl-4-propylpiperazine was
used as a solubilizing group (12k), the introduction of
2-morpholin-4-yl-acetamide or 2-piperazin-1-yl-acetam-
ide solubilizing chains (12l and 12m, respectively) was
detrimental for the activity. A dramatic difference in
activity between the two enantiomers was observed
when comparing 12e with 12f and 12g with 12h, where
the more active enantiomer was 35 times more potent
that the other. Epimerizing of the chiral center of posi-
tion 3 also deceases the IC50 value (12c and 12d). The
replacement of the iodine substitution of 12e by 1-pro-
pynyl moiety afforded 12j, which was found less potent.
Figure 4. The binding model of the Hdm2-12a complex. In addition to
the surface representation of the protein, Val93 is shown in stick form,
highlighting the proposed intermolecular hydrogen bond with the
aniline N of 12a (see text).
The primary evaluation of the inhibitory activity of final
compounds was assessed using a displacement assay of a
fluorescent peptide substrate mimicking the sequence of
the p53 interaction with Hdm2 (FP assay).6a Induction
of PIG3 production in JAR carcinoma cells, which are
known to overexpress both Hdm2 and p53, was used
as secondary screening to confirm that the cellular activ-
ity of 12a–m is mediated by p53 activation. The PIG3
production was measured by ELISA.6d The results of
the FP1 binding assay and the PIG3 production in cell
culture are reported in Table 1.
Table 2 shows the different activity that some of our
most potent compounds have in wild-type versus mu-
tant p53 tumor cell lines. As expected based on the
mechanism of action, these compounds showed good
selectivity for inhibiting growth of the wild-type p53
MCF7 tumor cell line and a concomitantly reduced
activity in the mutant p53 MDA MB 231 cell line.6d
Depending upon the p53 activation level, wt-p53 tu-
mor cell lines can be driving toward either a cytostatic

3118
J. J. Marugan et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3115–3120
Table 1. SAR using FP IC50 and PIG3 production
O
2
R
4
R
N
3
R
N
1
O
R
Compound
12aa
12b
R1
R2
R3
R4
FP IC50 (lM)
0.250
0.856
2.7
PIG-3
1.1 lM
3.3 lM
10 lM
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
O
O
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
I
47
74
79.3
3.5
196
OH
OH
NH
2
I
60
2.8
208.6
23.9
36.4
1404
103.7
97
O
O
O
O
12c
I
N
N
N
12d
I
6.25
37.3
153
32.9
683
84.1
58
12e
I
0.367
13.1
NH
2
12f
I
N
O
NH
2
12g
I
2.39
30
O
O
O
O
O
12h
I
>12.5
0.394
0.787
0.546
5.4
12i
I
96.4
54.8
84.0
31
338.1
265.0
396.9
32
458.85
474.1
412.2
24
NH
2
O
12j
N
NH
2
N
12k
I
I
I
N
NH
2
O
N
N
12l
O
NH
12m
1.55
8.6
23.3
33.3
NH
2
O
a Compound 12a is the racemic mixture.
or apoptotic event. In order to assess whether these
inhibitors can drive the cell toward apoptosis, and
therefore anticipate the possible response of our
antagonists on a tumor growth xenograft, we mea-
sured the caspase activity in JAR cells. Figure 5 shows
the rate of caspase-3 and caspase-7 activation in JAR

J. J. Marugan et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3115–3120
3119
Table 2. Comparison of the activity of compounds inhibiting growth
of wt-p53 tumor cells (MCF7) versus mutant p53 tumor cells (MDA-
MB-231)
nists. Structure-base predictions have provided reli-
able guidance in exploring this class of molecules.
Increased potency was obtained by the introduction
of an amine functional group in the ortho position
of the benzylic ring, leading to the formation of an
additional hydrogen bond with Val93. We have also
shown that the replacement of the valeric acid side
chain with different solubilizing groups has major im-
pact on the cellular potency of this class of mole-
cules. Additionally, these series of compounds
induce PIG3 production in cell-based assays, and dis-
play selectivity towards wild-type p53 tumor cells.
The caspase induction shown with compounds 12i,
12e, and 12b indicates that benzodiazepinediones are
capable to activate an apoptotic response in cell-
based assay.
O
R2
R4
N
Cl
N
O
R1
R1
R2
R4
BrdU IC50
(lM)
MCF
MDA
Cl
O
O
12a
12e
12i
I
I
I
16
80
10
57
10
8
OH
Cl
Cl
Cl
Cl
N
O
1.3
NH
NH
NH
2
2
2
2
References and notes
1. Oren, M. J. Biol. Chem. 1999, 274, 36031.
1.1
1.7
0.8
O
2. (a) Kubbutat, M. H. G.; Vousden, K. H. Mol. Med.
Today 1998, 4, 250; (b) Harris, C. C. J. Natl. Cancer Inst.
1996, 88, 1442; (c) Vogelstein, B.; Kinzler, K. W. Cell
1992, 70, 523.
3. (a) Picksley, S. M.; Lane, D. P. Bioessays 1993, 15, 689; (b)
Piette, J.; Nell, H.; Marechal, V. Oncogene 1995, 15, 1001;
(c) Gasco, M.; Crook, T. Drug Resist. Updat. 2004, 6, 323.
4. Momand, J.; Jung, D.; Wilczynski, S.; Niland, J. Nucleic
Acids Res. 1998, 26, 3453.
O
12j
N
N
12k
I
N
NH
5. (a) Garcia-Echeverria, C.; Chene, P.; Blommers, M. J. J.;
Furet, P. J. Med. Chem. 2000, 43, 3205; (b) Zhao, J.; Wang,
M.; Chen, J.; Luo, A.; Wang, X.; Wu, M.; Yin, D.; Liu, Z.
Cancer Lett. 2002, 183, 69; (c) 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. Science 2004, 303, 844.
Caspase 3/7 Induction
3
2
1
l2i
6. (a) Grasberger, B. L.; Lu, T.; 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.;
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. J. Med. Chem. 2005, 48, 909;
(b) Parks, D. J.; LaFrance, L. V.; Calvo, R. R.;
Milkiewicz, K. L.; Gupta, V.; Lattanze, J.; Ramachan-
dren, K.; Carver, T. E.; Petrella, E. C.; Cummings, M.
D.; Maguire, D.; Grasberger, B. L.; Lu, T. Bioorg. Med.
Chem. Lett. 2005, 15, 765; (c) Raboisson, P.; Marugan, J.
J.; Koblish, H. K.; Lu, T.; Zhao, S.; Player, M. R.;
Maroney, A. C.; Huebert, N. D.; Reed, R. L.; Cum-
mings, M. D. Bioorg. Med. Chem. Lett. 2005, 15, 1857;
(d) Koblish, H. K.; Zhao, S.; Franks, C. F.; Donatelli, R.
R.; LaFrance, L. V.; Leonard, K.; Gushue, J. M.; Parks,
D. J.; Calvo, R. R.; Milkiewicz, K. L.; Marugan, J. J.;
Raboisson, P.; Cummings, M. D.; Grasberger, B. L.; Lu,
T.; Molloy, C. J.; Maroney, A. C. Mol. Cancer Ther.
2006, 5, 160(e) Leonard, K.; Marugan, J.J.; Raboisson,
P.; Calvo, R.; Gushue, J.M.; Koblish, H.K.; Lattanze, J.;
Zhao, S.; Cummings, M.D.; Player, M.R.; Maroney, A.;
Lu, T. Bioorg. Med. Chem. Lett., in press.
l2e
l2b
DMSO
0
50 μm
25 μm
10 μm
Treatment
Figure 5. Caspase induction in JAR cells, 48-h exposure.
cells treated with some of our best antagonists.8 At 25
and 50 lM, the rate of caspase-3 and -7 production is
substantially increased compared to that observed for
cells treated with DMSO control. These data clearly
indicate that inhibition of the Hdm2 interaction with
p53 leads to activation of an apoptotic response in
JAR cells.
7. CCDC 601389 contains the supplementary crystallographic
data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
In summary, we have reported herein the synthesis of
enantiomerically pure 1,4-benzodiazepine-2,5-diones as
Hdm2 antagonists 12a–m as potent Hdm2 antago-

3120
J. J. Marugan et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3115–3120
8. Numerical data corresponding to Figure 5:
O
2
R
O
4
R
N
3
R
N
1
R
Compound
R1
R2
R3
R4
Fold
50 lM
25 lM
10 lM
Cl
O
O
2
5
Cl
Cl
Cl
I
I
I
1.97
1.37
2.87
0.97
OH
Cl
Cl
2.89
0.9
N
O
NH
NH
2
9
2.92
1
2.84
1
1.29
1
O
2
DMSO