Rational design, synthesis, and biological evaluation of rigid pyrrolidone analogues as potential inhibitors of prostate cancer cell growth

Article abstract of DOI:10.1016/S0960-894X(01)00097-X

In view of its role in tumor promotion and signal transduction, protein kinase C (PKC) has proven to be an exciting target for cancer therapy. With the aid of molecular modeling, we rationally designed and stereoselectively synthesized a new class of rigidified pyrrolidone-based PKC activators. Pyrrolidone 15 was found to exhibit reasonable affinity for PKCδ, with lower affinity for the other isozymes tested. Pyrrolidone 2 causes the dose-dependent induction of apoptosis in LNCaP prostate cancer cells. This apoptotic effect could be markedly potentiated by the use of LNCaP cells overexpressing the PKCα or δ isozymes.

Full text of DOI:10.1016/S0960-894X(01)00097-X

Bioorganic & Medicinal Chemistry Letters 11 (2001) 955–959
Rational Design, Synthesis, and Biological Evaluation of Rigid
Pyrrolidone Analogues as Potential Inhibitors of
Prostate Cancer Cell Growth
Lixin Qiao,a,y Lian-Yun Zhao,a Suo-Bao Rong,a Xiong-Wu Wu,a Shaomeng Wang,a
b
c
Teruhiko Fujii,b Marcelo G.Kazanietz, Laura Rauser,c Jason Savage,c Bryan L.Roth,
a,
Judith Flippen-Andersond and Alan P.Kozikowski *
aDrug Discovery Program, Department of Neurology, Georgetown University Medical Center, Washington, DC 20007, USA
bCenter for Experimental Therapeutics and Department of Pharmacology, University of Pennsylvania School of Medicine,
Philadelphia, PA 19104-6160, USA
cNIMH Psychoactive Drug Screening Program, Department of Biochemistry and Neurosciences,
Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 44106, USA
dNaval Research Laboratory, Washington, DC 20375, USA
Received 20 October 2000; accepted 23 January 2001
Abstract—In view of its role in tumor promotion and signal transduction, protein kinase C (PKC) has proven to be an exciting
target for cancer therapy.With the aid of molecular modeling, we rationally designed and stereoselectively synthesized a new class
of rigidified pyrrolidone-based PKC activators.Pyrrolidone 15 was found to exhibit reasonable affinity for PKCd, with lower
affinity for the other isozymes tested.Pyrrolidone 2 causes the dose-dependent induction of apoptosis in LNCaP prostate cancer
cells.This apoptotic effect could be markedly potentiated by the use of LNCaP cells overexpressing the PKC a or d isozymes.
# 2001 Elsevier Science Ltd.All rights reserved.
6
Cancer is the second leading cause of death in the Uni-
ted States after cardiovascular disease, and it is pro-
jected that cancer will become the leading cause of death
within the coming years.1 After lung cancer, prostate
cancer is the second most fatal cancer for American
men.As an important signal transducing enzyme that is
crucial for all aspects of cellular development, differ-
entiation, and transformation,2 PKC has been proposed
to play an important role in carcinogenesis, metastasis,3
and chemotherapy-associated multidrug resistance.4
The PKC activator 12-O-tetradecanoylphorbol-13-ace-
tate (TPA) functions differently in androgen-sensitive
and androgen-independent human prostate cancer
cells.5 TPA activates cell death in androgen-sensitive
LNCaP cells, in contrast to androgen-independent DU-
145 or PC-3 cells.This inhibitory effect in androgen-
sensitive cells was significantly decreased by the compe-
titive PKC inhibitor, staurosporine.A recent review
summarizes other lines of evidence that the expression
of constitutive PKCa activity is required for the survival
and growth of androgen-independent human prostate
cancer cells.Moreover, systemic delivery of PKC a anti-
sense oligonucleotides (ISIS 3521) has been shown to
result in the marked growth inhibition of U-87 xeno-
grafts without affecting the levels of other PKC iso-
forms.7 Taken together, these studies imply that PKCa
may serve as a novel target in the treatment of prostate
cancer.In this report, we describe our results with a
rigidified pyrrolidone-based PKC activator 2, which is
shown to cause the dose-dependent induction of apop-
tosis in LNCaP prostate cancer cells.This apoptotic
effect is shown to be potentiated in LNCaP cells over-
expressing the PKCa or d isozymes.
The PKC gene family consists presently of 11 genes,
which are divided into four subgroups.Each PKC iso-
zyme contains a regulatory domain and a catalytic
domain.Within the regulatory domain, both classical
(a, b, and g) and novel (d, e, Z, and y) PKC isozymes
*Corresponding author.Te:l. +1-202-687-0686; fax: +1-202-687-
5065; e-mail: kozikowa@giccs.georgetown.edu
yCurrent address: NeuroLogic Incorporated, 15010 Broschart Road,
Suite 200, Rockville, MD 20850, USA.
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd.All rights reserved.
PII: S0960-894X(01)00097-X

956
L. Qiao et al. / Bioorg. Med. Chem. Lett. 11 (2001) 955–959
contain two cysteine-rich zinc finger domains (C1A and
C1B), which constitute the diacylglycerol/phorbol ester
binding site.In 1995, the X-ray crystal structures of
PKCd C1B in complex with phorbol 13-acetate revealed
precisely how phorbol ester binds to PKC8 and pro-
vided a solid structural basis for the rational design of
new PKC modulators.9,10 Based upon the structure of
the potent PKC activator 8-decynylbenzolactam (1),
and with the aid of molecular modeling, we designed a
series of pyrrolidone analogues as a new class of PKC
modulators.11 These pyrrolidone analogues maintain
the hydrogen bonding network and hydrophobic inter-
actions considered crucial for interaction with PKC,
and were found to possess reasonably good affinity for
PKC and, in a few cases, some isozyme selectivity.
Upon taking into consideration both the rotational
flexibility of the aromatic ring in these pyrrolidone ana-
logues together with the structure of the more potent
benzolactams, we designed the rigidified pyrrolidone
structures with the aim to possibly reach comparable
potency by reducing conformational mobility.Thus, the
benzolactam N-1 atom was replaced by a carbon atom,
followed by shifting the isopropyl group from the pre-
vious C-2 position to the new carbon.C-2 and C-6 were
subsequently joined to give the pyrrolidone system.
Finally, the stereochemistry of all chiral centers was
adjusted based upon results derived from molecular
modeling to arrive at the structure of pyrrolidone 2
shown in Figure 1.After energy minimization, the
resulting conformation of compound 2 was super-
imposed upon the benzolactam 1.As depicted in Figure 2,
the crucial hydroxyl group and amide residue are pre-
served in compound 2, and these groups engage in a
strong hydrogen bonding network with PKCd C1B.The
front methyl group on the marked chiral center mimics
the N–Me in benzolactam 1, interacting with Leu250 of
PKC (Fig.3).Similarly, the rear isopropyl group of
pyrrolidone 2 interacts with the side chain of Leu254,
thereby mimicking the isopropyl group of benzolactam
1.The same orientation of the phenyl group in com-
pound 2 allows for strong hydrophobic interactions
with Pro241 of PKC.
Retrosynthetic analysis revealed that pyrrolidone 2 could
be easily assembled via conjugate addition and radical
cyclization as the key steps.The bromide 6 required for
preparation of the corresponding Grignard reagent is
readily available from 3-hydroxybenzaldehyde.Bromi-
nation of 3-hydroxybenzaldehyde (3) with bromine in
dichloromethane gave rise to 2-bromo-5-hydroxy-
benzaldehyde (4) in quantitative yield.Subsequent
nucleophilic addition of 2 equiv of t-butylmagnesium
chloride provided the crystalline benzyl alcohol 5 in
good yield.Next, a cationic rearrangement induced by
Figure 3. Superimposition of the parent structure of pyrrolidone 2
upon the benzolactam 1 in PKCd C1B.
Figure 1. Design concept of pyrrolidone 2.
Figure 2. Overall features of the binding model for pyrrolidone 2 and benzolactam 1 in complex with PKCd C1B.

L. Qiao et al. / Bioorg. Med. Chem. Lett. 11 (2001) 955–959
957
boron trifluoride etherate followed by O-benzylation
furnished the desired bromide 6 bearing a tetrasubstituted
olefin.Copper(I)-catalyzed conjugate addition of
Grignard reagent derived from bromide 6 to lactam 7
furnished compound 8 stereoselectively.The steric hin-
drance provided by the tetrasubstituted olefin prevents
free rotation of the phenyl ring.Hence, lactam 8 was
observed to exist as a pair of conformers by NMR,
presumably with the olefin located on one side or the
other of the pyrrolidone ring.Introduction of the ring
constraint, which involves creation of a new five-mem-
bered ring, was then accomplished through radical
cyclization initiated by tin hydride reduction of a phe-
nylselenyl group introduced a to the carbonyl group.
The cyclized products 9 and 10, and the recovered pro-
duct 8 were separated by careful column chromato-
graphy on silica gel.The stereochemistry of lactams 9
and 10 were assigned by 1H–1H COSY and 1H–1H
NOESY as well as by X-ray diffraction of lactam 10.
Following the general procedures reported previously,11
intermediate 9 was then readily transformed into the
target molecule, pyrrolidone 2 (Scheme 1).Compound
13, which contains an ester function in place of the decy-
nyl appendage of 2, was also prepared from intermediate
9.For purposes of comparison, compound 15, the dia-
stereoisomer of compound 13, was synthesized in a
similar manner from intermediate 10 (Scheme 2).
Pyrrolidones 2, 13, and 15, as well as their progenitor,
benzolactam 1, were evaluated for their ability to dis-
place phorbol 12,13-dibutyrate (PDBu) binding from
five recombinant PKC isozymes, namely, PKCa, PKCb,
PKCg, PKCd, and PKCe (Table 1).12 Compared to the
benzolactam 1, these three new rigid pyrrolidones
exhibited 5- to 30-fold lower potency in binding to the
classical and novel PKC isozymes.These results suggest
that some important hydrophobic interactions that exist
between the benzolactam 1 and PKC are likely to have
been lost upon conversion to the rigid pyrrolidone
structures.The methyl group and the isopropyl group at
the starred chiral center of pyrrolidone 2 (Fig.1) may
thus only partially mimic the N-methyl and isopropyl
groups in benzolactam 1.Interestingly, the PKC
binding affinity of pyrrolidone 15 is almost identical to
d
Scheme 1. Synthetic route to pyrrolidone 2.

958
L. Qiao et al. / Bioorg. Med. Chem. Lett. 11 (2001) 955–959
Scheme 2. Preparation of pyrrolidones 13 and 15 from intermediates 9 and 10, respectively.
that of the benzolactam 1, and it is greater than that
found for compounds 2 and 13.This result indicates
that the gain in binding affinity stemming from the
hydrophobic interaction of the isopropyl group in com-
pound 15 with Pro241 of PKCd may override the loss in
the interaction with Leu250 and Leu254 of PKCd (Fig.3).
Accordingly, it may be possible to further modify this
particular chiral center in pyrrolidone 2 to enhance such
hydrophobic interactions with PKC.
tions of pyrrolidone 2 or PMA for 1 h.Cells were
collected 24 h later and stained with DAPI.The inci-
dence of apoptosis in each preparation was analyzed by
counting 500 cells and determining the percentage of
apoptotic cells.Results are the mean ꢀstandard error of
three independent experiments.
It is well established that PKCa and PKCd are the iso-
zymes that mediate phorbol ester-induced apoptosis in
LNCaP cells.13cꢁe A model that had been used to assess
isozyme-specificity is the overexpression of specific
PKCs by an adenoviral system.Infection of LNCaP
cells with adenoviruses for PKCs leads to a high
expression of the corresponding PKC isozyme, as
determined by Western-blot analysis, kinase activity, or
[3H]PDBu binding.13e Using this system, we found a
marked potentiation of PMA-induced apoptosis in
LNCaP cells by overexpression of PKC isozymes,13e as
also shown in Figure 4B.Remarkably, a similar effect
was observed with pyrrolidone 2, where overexpression
of either PKCa or PKCd markedly potentiates the
apoptotic response in LNCaP prostate cancer cells (Fig.
4A).As expected, a control adenovirus (LacZ adeno-
virus) did not affect the apoptotic effect of either pyrro-
lidone 2 or PMA.
Since the chemical yield of intermediate 9 is better than
that of its diastereoisomer 10, we scaled up the synthesis
of pyrrolidone 2 to allow for further biological studies.
Pyrrolidone 2 is also preferable over analogues 13 and
15, as the latter may be subject to the action of estera-
ses, resulting in cleavage of the hydrophobic side-chain
residue.The capability of inducing apoptosis by pyrro-
lidone 2 was evaluated in a model of LNCaP prostate
cancer cells.It is well established that in this model
phorbol esters (e.g., PMA) activate PKC to induce
apoptosis.13 LNCaP cells were treated with pyrrolidone
2 for 1 h, and apoptosis was assessed 24 h later.As
depicted in Figure 4A, pyrrolidone 2 (1–30 mM) induces
a dose-dependent apoptotic effect in LNCaP cells.For
comparison, a representative experiment using the
phorbol ester PMA is shown in Figure 4B.
The present efforts reveal that modification of the par-
ent benzolactam 1 to produce the rigid pyrrolidone
analogues leads to pyrrolidone 15 that shows compar-
able binding affinity at PKCd, while at the same time
exhibiting reduced affinity for the other isozymes tested.
This result provides structural information for improv-
ing PKCd isozyme selectivity.Moreover, while the pyr-
rolidone 2 was found to be less potent than PMA at
binding to PKC, it still induces levels of apoptosis in
LNCaP cells overexpressing either the PKCa or d iso-
zymes.As other evidence indicates that lower affinity
ligands may prove to be safer when administered systemi-
cally, the present results may be relevant to anticancer
drug development.
Control cells or cells infected for 14 h with adenoviruses
for LacZ, PKCa, or PKCd (multiplicity of infec-
tion=10 pfu/cell) were treated with different concentra-
Table 1. Ki values (nM) for the inhibition of [3H]PDBu binding to
human recombinant PKC isozymes by the compounds tested
Compound PKCa
PKCb
PKCg
43ꢀ7
PKC d
PKCe
1
2
23ꢀ4
70ꢀ18
114ꢀ37
40ꢀ14
580ꢀ214 701ꢀ217 279ꢀ50 1630ꢀ391 469ꢀ140
13
15
PMA
395ꢀ71
674ꢀ249 1052ꢀ284 327ꢀ65
2.5ꢀ0.5 1.8ꢀ0.6 4.9ꢀ0.9
593ꢀ178 209ꢀ38 1370ꢀ329 515ꢀ155
150ꢀ64 228ꢀ43
0.8ꢀ0.2 0.45ꢀ0.15

L. Qiao et al. / Bioorg. Med. Chem. Lett. 11 (2001) 955–959
959
Figure 4. Induction of apoptosis in LNCaP prostate cancer cells by pyrrolidone 2 (Panel A) or PMA (Panel B).
Acknowledgements
7. (a) Dean, N. M.; McKay, R.; Condon, T.; Bennett, C. F. J.
Biol. Chem. 1994, 269, 16416.(b) Dean, N.M;. McKay, R.
Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 11762.(c) Dean, N.M.;
McKay, R.; Miraglia, L.; Howard, R.; Cooper, S.; Giddings,
J.; Nicklin, P.; Meister, L.; Ziel, R.; Geiger, T.; Muller, M.;
Fabbro, D. Cancer Res. 1996, 56, 3499.
8. Zhang, G.; Kazanietz, M. G.; Blumberg, P. M.; Hurley,
J.H. Cell 1995, 81, 917.
9. Wang, S.; Liu, M.; Lewin, N. E.; Lorenzo, P. S.; Bhatta-
charrya, D.; Qiao, L.; Kozikowski, A. P.; Milne, G. W. A.;
Blumberg, P.M. J. Med. Chem. 1999, 42, 3436.
This work was supported by NIH grant CA79601 (to A.
P.K.), the Department of Defense DMAD 17-93-V-
3018 (to A.P.K.), the NIMH Psychoactive Drug
Screening Program (MH 01366 to B.L.R.), the Depart-
ment of Defense Prostate Cancer Research Program (to
M.G.K.), and in part by NIDA and the Office of Naval
Research (to J.F.-A.).
10. Kozikowski, A. P.; Wang, S.; Ma, D.; Yao, J.; Ahmad, S.;
Glazer, R. I.; Bogi, I.; Acs, P.; Modarris, S.; Blumberg, P. M.
J. Med. Chem. 1997, 40, 1316.
References
11. Qiao, L.; Wang, S.; George, C.; Lewin, N. E.; Blumberg,
P.M;. Kozikowski, A.P. J. Am. Chem. Soc. 1998, 120, 6629.
12. Roth, B. L.; Mehegan, J. P.; Jacobwitz, D. M.; Robey, F.;
Iadarola, M.J. J. Neurochem. 1989, 52, 215.
1.Gibbs, J.B. Science 2000, 287, 1969.
2.(a) Goekjian, P.G;. Jirousek, M.R.
1999, 6, 877.(b) Mochly-Rosen, D;. Kauvar, L.M.
Curr. Med. Chem.
Adv.
Pharmacol. 1998, 44, 91.(c) Newton, A.C. J. Biol. Chem.
1995, 270, 28495.(d) Glazer, R.I. Curr. Pharm. Des. 1998, 4,
277.
3.Herbert, J.M. Biochem. Pharmacol. 1993, 45, 527.
4. Blobe, G. C.; Sachs, C. W.; Khan, W. A.; Fabbro, D.;
Stabel, S.; Westal, W. C.; Obeid, L. M.; Fine, R. L.; Hannun,
U.A. J. Biol. Chem. 1993, 268, 658.
13. (a) Day, M. L.; Zhao, X.; Wu, S.; Swanson, P. E.; Hum-
phrey, P.A. Cell Growth Diff. 1994, 5, 735.(b) Garzotto, M;.
White-Jones, M.; Jiang, Y.; Ehleiter, D.; Liao, W.-C.; Hai-
movitz-Friedman, A.; Fuks, Z.; Kolesnick, R. Cancer Res.
1998, 58, 2260. (c) Powell, C. T.; Brittis, N. J.; Stec, D.; Hug,
H.; Heston, W. D. W.; Fair, W. R. Cell Growth Diff. 1996, 7,
419. (d) Gschwend, J. E.; Fair, W. R.; Powell, C. T. Mol.
´
5.Henttu, P;. Vihko, P.
1998, 244, 167.
6.O’Brian, C.A. Oncol. Rep. 1998, 5, 305.
Biochem. Biophys. Res. Commun.
Pharmacol. 2000, 57, 1224.(e) Fujii, T.; Garcı a-Bermejo, M.L.;
´
Bernabo, J. L.; Caamano, J.; Ohba, M.; Kuroki, T.; Li, L.;
Yuspa, S.H.; Kazanietz, M.G. J. Biol. Chem. 2000, 275, 7574.