Aryl indanyl ketones: Efficient inhibitors of the human peptidyl prolyl cis/trans isomerase pin1

Article abstract of DOI:10.1002/anie.200601569

A good imitation: Aryl-1-indanyl ketones are found to be highly efficient, reversible, cell-penetrating inhibitors of the human peptidyl prolyl cis/trans isomerase Pin1. Owing to their structure 1, they are assumed to mimic the transition state 2 of the e

Full text of DOI:10.1002/anie.200601569

Communications
Enzyme Inhibitors
DOI: 10.1002/anie.200601569
Aryl Indanyl Ketones: Efficient Inhibitors of the
Human Peptidyl Prolyl cis/trans Isomerase Pin1**
Sebastian Daum, Frank Erdmann, Gunter Fischer,
Boris FØaux de Lacroix, Anahita Hessamian-Alinejad,
Sabine Houben, Walter Frank, and Manfred Braun*
The rigidity of the amide bond is a feature that distinctly
influences the structure of proteins. The rotation about the
imidic peptide bond preceding proline has to surmount an
activation barrier of 75 to 100 kJmolꢀ1 for the conversion of
the cis-conformer 1 into trans-3 (Scheme 1). Thereby, the
torsion angle w changes from 0 to 1808. In unfolded proteins,
the trans-isomer 3 predominates; however, native proteins
frequently contain a specific prolyl bond in the cis conforma-
tion 1.[1]
Scheme 1. Cis–trans isomerization of peptidyl–prolyl bonds. Aryl 1-
indanyl ketones as mimics of the twisted-amide transition state 2.
Peptidyl prolyl cis/trans isomerases (PPIases)are enzymes
that catalyze the cis–trans interconversion of prolyl bonds
(1Ð3).[2] Discovered in 1984, they have been found to be an
important group of enzymes with a strong impact on the slow
processes involved in protein folding.[3] Concerning the
mechanism of the enzymatic catalysis exhibited by PPIases,
various hypotheses have been developed to rationalize how
[*] Dipl.-Chem. B. FØaux de Lacroix, Dr. A. Hessamian-Alinejad,
S. Houben, Prof. Dr. W. Frank, Prof. Dr. M. Braun
Wissenschaftliche Einrichtung Chemie
Universität Düsseldorf
Universitätsstrasse 1, 40225 Düsseldorf (Germany)
Fax: (+49)211-811-5079
E-mail: braunm@uni-duesseldorf.de
S. Daum, Dr. F. Erdmann, Prof. Dr. G. Fischer
Max-Planck-Forschungsstelle für Enzymologie der Proteinfaltung
Weinbergweg 22, 06120 Halle (Germany)
[**] This workwas supported by the Deutsche Forschungsgemeinschaft
(Br 604/14-1,2).
7454
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7454 –7458

Angewandte
Chemie
the energy of the transition state 2 in the uncatalyzed
isomerization of cis-1 and trans-3 is lowered to the exper-
imentally measured values of the enzyme catalyzed reac-
tion.[1,2b]
easily available by coupling lithiated arenes 5a,b, generated
by a bromine–lithium exchange from bromoarenes 4a,b, with
monodeprotonated indanecarboxylic acid 6a.[12] As an alter-
native to the lithium salt 6b, the Weinreb-type[13] amide 6c
was also found to react readily with aryllithium 5 to deliver
the ketones 7. Generally, this variant provided higher yields.
The cleavage of the methyl ether group in compounds 7a,b
occurred upon treatment with aluminum chloride[14] to give
the phenols 8a,b. For the introduction of a methyl substituent
in the a-carbonyl position, the enolate generated from ketone
7b was treated with methyl iodide to give racemic product 9.
A novel domino reaction led to the biaryl indanyl ketone
10, as shown in Scheme 3. When two equivalents of aryl-
Among the three families of human PPIases, the parvulin-
like Pin1 is unique in its substrate specificity which has made
it possible to specifically accelerate the cis–trans isomeriza-
tion of PO3H2Ser(PO3H2Thr)–Pro bonds in short peptides
and proteins.[4] Therefore, Pin1 interacts with and regulates
many key phosphoproteins of the eukaryotic cell cycle, such
as cyclin D1, p53, or b-catenin.[5] In several human cancer cell
lines, the depletion of Pin1 causes cell cycle arrest. Moreover,
the enzyme has been found to be overexpressed in many
tumor cell lines.[6] Pin1 has also been found to be involved in
Alzheimerꢀs disease.[7] To date, however, few compounds
have been reported to act as inhibitors of Pin1. Among them,
the natural naphthoquinone juglone inactivated Pin1 in an
irreversible manner by minor structural alterations at the
active site after the Michael addition of a thiol group.[8]
Several polycyclic aromatic compounds[9] and natural peptide
mimetics[10] have also been reported to inhibit Pin1.
In our search for efficient reversible inhibitors of human
Pin1 (hPin1), we were guided by the idea that the “twisted-
amide” transition-state model 2 along with the concomitant
change of hybridization about the ring nitrogen atom might
be readily mimicked by aryl 1-indanyl ketones, because we
anticipated them to have a twisted conformation in front of
the five-membered ring. Its similarity to the “twisted amide” 2
is indicated in Scheme 1.
Despite their relatively simple structure, only a few
compounds containing the aryl 1-indanyl-ketone moiety
have been synthesized.[11] Most of them served as intermedi-
ates for fredericamycin A[11c–g] and, in only a few cases, were
nonracemic products synthesized.[11d,i] Thus, general routes to
aryl indanyl ketones had to be developed to make the
preparation of specifically substituted products possible. As
shown in Scheme 2, monoaryl indanyl ketones 7a,b were
Scheme 3. Domino reaction between aryllithium 5b and carboxylate
6b. Synthesis of ketones 11 and 12 derived from 10. Reagents and
conditions: a) nBuLi, Et2O, ꢀ788C; b) ꢀ788C to reflux, 27%; c) AlCl3,
CH2Cl2, 08C to room temperature, 59%; d) BBr3, CH2Cl2, 71%;
e) HNO3, HOAc, 108C, 79%; f) H2/Pd/C, THF, 73%; g) biotinyl
chloride, pyridine, 8%; h) H2/Pd/C, EtOH, 95%.
lithium 5b, generated by bromine–lithium exchange from 4b,
were allowed to react with the lithium salt 6b of indane
carboxylic acid, the biaryl indanyl ketone 10 resulted as the
only product apart from protonated starting material 5b (H
instead of Li)and carboxylic acid 6a. Clearly, not only the
addition to the carboxylate 6b had occurred, but also a
substitution of the fluorine atom by the second equivalent of
the aryllithium 5b serving as a nucleophile. Although the
Scheme 2. Synthesis of monoaryl indanyl ketones 7–9. Reagents and
conditions: a) nBuLi, Et2O, ꢀ788C; b) nBuLi, Et2O, ꢀ78 to ꢀ108C;
c) 1. SOCl2, reflux; 2. HN(OMe)Me·HCl, pyridine, 08C to room temper-
ature; d) AlCl3, CH2Cl2, 08C; e) 1. LiNiPr2, THF, ꢀ78 to 08C, 2. MeI,
ꢀ208C to room temperature.
Angew. Chem. Int. Ed. 2006, 45, 7454 –7458ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7455

Communications
Table 1: Inhibition constants of aryl indanyl ketones 7–9, 11, and 12.
question remains open whether this step follows an addition–
elimination or an aryne pathway, the complete regiocontrol is
remarkable. The domino reaction does not occur when,
instead of the carboxylate 6b, the amide 6c is treated with two
equivalents of 5b.
An unambiguous confirmation of the structure of ketone
10, which forms as a single isomer, comes from a crystal-
structure analysis of phenolic ketone 11a (Figure 1),[15] the
Entry
Ketone
Ki [mm]
Entry
Ketone
Ki [mm]
1
2
3
4
5
6
7
8
7a
7b
8a
8b
9
11a
11b
11c
6.2ꢁ0.4
8.7ꢁ0.2
5.1ꢁ0.8
15.0ꢁ3.0
19.0ꢁ4.9
1.7ꢁ0.2
1.1ꢁ0.2
4.9ꢁ0.7
9
10
11
12
13
14
15
11d
12a
12b
12c
12d
(R)-9
(S)-9
1.4ꢁ0.9
3.1ꢁ0.6
0.5ꢁ0.1
0.2ꢁ0.1
0.4ꢁ0.1
5.6ꢁ0.9
51.0ꢁ9.0
(entries 1–10)have Ki values in the micromolar range. The
hydrogen-bonded hydroxy group adjacent to the carbonyl
group is not crucial to Pin1 inhibition, although the phenolic
derivatives are superior to the phenol ethers. A significant
improvement in Pin1 inhibition originates from the introduc-
tion of nitro groups. Thus, mono- and dinitro-substituted
ketones 12b and 12c (Table 1; entries 11 and 12)led to a
strong inhibition with Ki values in the sub-micromolar range.
Remarkably, the replacement of a nitro group in 12b by an
amino substituent in 12d did not influence Pin1 inhibition to a
major extent (Table 1; entry 13).
Ketone 11d with a biotin side chain allowed the reversi-
bility of Pin1 inhibition to be demonstrated. For this experi-
ment the activity assay was performed at a fixed concen-
tration of 11d, leading to 80% inhibition of the Pin1 PPIase
activity. The addition of increasing concentrations of strepta-
vidin (a stronger competitive binder for 11d)resulted in the
full return of Pin1 activity whereas the addition of strepta-
vidin presaturated with biotin had no effect (Figure 2). This
Figure 1. Molecular structure of ketone 11a. Thermal ellipsoids are set
at 30% probability; radii of hydrogen atoms are chosen arbitrarily.
Selected interatomic distances [Š] and angles [8]: C1–O1 1.2443(15),
C10–O2 1.3570(15), O2–H1 0.996(18), H1···O1 1.639(18), O1···O2
2.5570(15); O2-H1···O1 151.2(15), C3-C2-C8A 101.55(12); O1-C1-C9-
C10 5.01(18), O1-C1-C2-C8A 92.15(15), O1-C1-C2-C3 ꢀ20.98(19), C2-
C1-C9-C10 ꢀ174.96(12), C14-C13-C15-C16 57.77(17).
product of a selective demethylation of the methoxy sub-
stituent in the position ortho to the carbonyl group. Although
the yield of the domino reaction is moderate (25–30%), it
opens an easy way to the ketone 10 that in turn serves as a
starting material for further derivatives. Thus, a regioselective
ortho-nitration of the phenol 11a was possible and gave the
product 11b as a single isomer. Catalytic hydrogenation
afforded the aromatic amine 11c that was coupled with
biotin-derived acid chloride to give the amide 11d. A
complete demethylation of biaryl indanyl ketone 10 by
means of boron tribromide[14] led to the formation of
bisphenolic ketone 12a. Upon treatment with nitric acid in
acetic acid, a mixture of nitro compounds 12b and 12c was
obtained. After separation, the former was reduced to the
aromatic amine 12d.
Whereas all the aryl indanyl ketones described herein
were obtained as racemic mixtures, the a-methyl-substituted
ketone 9 was chosen to be prepared in an enantiomerically
pure form. For this purpose, enantiomeric (R)- and (S)-1-
methyl-1-indanecarboxylic acids[16,17] served as starting mate-
rials. Thus, reaction of the corresponding Weinreb-type
amides with 5b delivered (R)- and (S)-9, respectively, in 45
to 48% overall yield.
Figure 2. Reversibility of Pin1/11d complex formation; PPIase activity
of 5 nm Pin1 in the absence and presence of 5 mm 11d, 10 mm
Streptavidin, 30 mm Biotin.
behavior indicates that the inhibitor is noncovalently attached
to Pin1 and the concentration-dependent formation of
inactive enzyme is not due to irreversible Pin1 denaturation
and aggregation.[10c]
The inhibition of hPin1 by the aryl indanyl ketones
prepared as described above was determined in a protease-
free PPIase assay with Suc-Ala-Glu-Pro-Phe-pNA as the
substrate.[18] The compounds exhibit substantial inhibition of
hPin1. As shown in Table 1, most of the aryl indanyl ketones
As expected, a substantial difference in the inhibition of
the enzyme occurred when the two enantiomers of ketone 9
were tested, and the corresponding Ki values differed by an
order of magnitude from each other (Table 1; entries 14 and
15). The different behavior of the enantiomers underlines the
7456
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7454 –7458

Angewandte
Chemie
idea of the similarity between the structure of aryl indanyl
ketones and the “twisted-amide” transition state, because the
binding sites in the transition state of the active center must
have a stereochemical component. More support for the
hypothesis comes from the crystal structure of ketone 11a
(Figure 1). The torsion angle between the aromatic-carbonyl
and the indanyl moiety (O1-C1-C2-C8A)of 92 8 is in accord
with the features of the twisted amide 2. In addition,
homologues of aryl indanyl ketones with four- or six-
membered rings instead of a five-membered one do not
display substantial Pin1 inhibition, a fact that supports the
“twisted-amide” hypothesis (data not shown).
incubated with the compounds and the intracellular b-catenin
level was determined by Western blotting with a specific anti-
b-catenin antibody and subsequent densitometric analysis
(Figure 4). The application of (R)-9 and rac-9 leads to a
decreased b-catenin content in the cells, whereas (S)-9 has
only a minor effect.
Next we asked whether application of selected com-
pounds of Table 1 would give rise to the same effects in cells
as those obtained after the depletion of Pin1 in cells by
genetic means. A luciferase reporter gene assay with a p53
response element in MCF-7 cells (breast cancer cell line)was
performed. In cells, the phosphorylated form of tumor-
suppressor protein p53 interacts with endogenous Pin1 after
DNA damage by chemotherapeutic drugs or irradiation. In
this pathway, Pin1 induces conformational changes which lead
to increased transactivation activity and higher proteolytic
stability of p53.[19]
As shown in Figure 3, the compounds (S)-9, (R)-9, and
rac-9 are able to reduce the stimulated activity of the p53
reporter gene in etoposid treated MCF-7 cells in a concen-
Figure 4. (S)-9, (R)-9, and rac-9 decrease the intracellular b-catenin
level in SH-SY5Y cells by Pin1 inhibition.
In summary, a new class of efficient, cell-penetrating,
reversible inhibitors of human Pin1 has been developed that
are based on the structural motif of aryl 1-indanyl ketones.
They not only display inhibition constants in sub-micromolar
range, but have shown biological activities and seem to be
promising candidates for the development of anticancer
drugs.
Received: April 20, 2006
Revised: June 22, 2006
Published online: October 18, 2006
!
*
*
Figure 3. Influence of (R)-9 ( ), (S)-9 ( ), and rac-9 ( ) on p53
reporter gene activity in etoposid-treated MCF-7 cells. Data are mean
values of three independent measurements (ꢁstandard deviation).
Keywords: antitumor agents · biological activity · chirality ·
.
enzymes · ketones
tration-dependent manner and according to their inhibitory
potencies. Similarly, cotransfection of Pin1-negative mouse-
embryo fibroblasts with the expression construct for an
inactive Pin1 variant failed to increase the stability and
transactivation activity of p53.[19c]
[1] Review: J. Fanghänel, Angew. Chem. 2003, 115, 506; Angew.
Chem. Int. Ed. 2003, 42, 490.
[2] Review: a)G. Fischer, Angew. Chem. 1994, 106, 1479; Angew.
Chem. Int. Ed. Engl. 1994, 33, 1415; b)G. Fischer, Chem. Soc.
Rev. 2000, 29, 119.
[3] G. Fischer, H. Bang, C. Mech, Biomed. Biochim. Acta 1984, 43,
1101.
In a second experiment, we investigated the effect of (S)-
9, (R)-9 and rac-9 on the oncogenic transcriptional activator
b-catenin. Several studies could demonstrate that Pin1
regulates the b-catenin turnover in tumor cells. Moreover,
the b-catenin level directly correlates with the PPIase activity
of Pin1, as shown by using a set of point-mutated Pin1
variants.[20] Cells of the tumor cell line SH-SY5Y were
[4] a)M. B. Yaffe, M. Schutkowski, M. Shen, X. Z. Zhou, P. T.
Stukenberg, J.-U. Rahfeld, J. Xu, J. Kuang, M. W. Kirschner, G.
Fischer, L. C. Cantley, K. P. Lu, Science 1997, 278, 1957; b)R.
Ranganathan, K. P. Lu, T. Hunter, J. P. Noel, Cell 1997, 89, 875;
c)J. D. Joseph, E. S. Yeh, K. I. Swenson, A. R. Means, E.
Winkler, Prog. Cell Cycle Res. 2003, 5, 477.
Angew. Chem. Int. Ed. 2006, 45, 7454 –7458ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7457

Communications
[5] A. Ryo, Y.-C. Liou, K. P. Lu, G. Wulf, J. Cell Sci. 2003, 116, 773.
[6] G. M. Wulf, A. Ryo, G. G. Wulf, S. W. Lee, T. Niu, V. Petkova,
K. P. Lu, Embo J. 2001, 20, 3459.
[7] a)Y.-C. Liou, A. Sun, A. Ryo, X. Z. Zhou, Z.-X. Yu, H.-K.
Huang, T. Uchida, R. Bronson, G. Bing, X. Li, T. Hunter, K. P.
Lu, Nature 2003, 424, 556; b)P. J. Lu, G. Wulf, X. Z. Zhou, P.
Davies, K. P. Lu, Nature 1999, 399, 784.
[8] L. Henning, C. Christner, M. Kipping, B. Schelbert, K. P.
Rücknagel, S. Grabley, G. Küllertz, G. Fischer, Biochemistry
1998, 37, 5953.
[19] a)P. Zacchi, M. Gostissa, T. Uchida, C. Salvagno, F. Avolio, S.
Volinia, Z. Ronai, G. Blandino, C. Schneider, G. Del Sal, Nature
2002, 419, 853; b)H. Zheng, H. You, X. Z. Zhou, S. A. Murray,
T. Uchida, G. Wulf, L. Gu, X. Tang, K. P. Lu, Z.-X. J. Xiao,
Nature 2002, 419, 849; c)G. M. Wulf, Y.-C. Liou, A. Ryo, S. W.
Lee, K. P. Lu, J. Biol. Chem. 2002, 277, 47976.
[20] A. Ryo, M. Nakamura, G. Wulf, Y.-C. Liou, K. P. Lu, Nat. Cell
Biol. 2001, 3, 793.
[9] T. Uchida, M. Takamiya, M. Takahashi, H. Miyashita, H. Ikeda,
T. Terada, Y. Matsuo, M. Shirouzu, S. Yokoyama, F. Fujimori, T.
Hunter, Chem. Biol. 2003, 10, 15.
[10] a)H. C. Wang, K. Kim, R. Bakhtiar, J. P. Germanas, J. Med.
Chem. 2001, 44, 2593; b)Y. Zhang, Mini-Rev. Org. Chem. 2004,
1, 359; c)E. Bayer, M. Thutewohl, C. Christner, T. Tradler, F.
Osterkamp, H. Waldmann, P. Bayer, Chem. Commun. 2005, 516;
d)D. Wildemann, F. Erdmann, B. Hernandez Alvarez, G.
Stoller, X. Z. Zhou, J. Fanghänel, M. Schutkowski, K. P. Lu, G.
Fischer, J. Med. Chem. 2006, 49, 2147.
[11] a)J. K. Stille, R. A. Newsom, J. Org. Chem. 1961, 26, 1375;
b)J. F. Bunnett, J. A. Skorcz, J. Org. Chem. 1962, 27, 3836;
c)D. L. J. Clive, A. G. Angoh, S. M. Bennett, J. Org. Chem. 1987,
52, 1339; d)W. Trypke, A. Steigel, M. Braun, Synlett 1992, 827;
e)D. L. J. Clive, Y. Tao, A. Khodabocus, Y.-J. Wu, A. G. Angoh,
S. M. Bennett, C. N. Boddy, L. Bordeleau, D. Kellner, G.
Kleiner, D. S. Middleton, C. J. Nichols, S. R. Richardson, P. G.
Vernon, J. Am. Chem. Soc. 1994, 116, 11275; f)D. L. Boger, I. C.
Jacobson, J. Org. Chem. 1990, 55, 1919; D. L. Boger, I. C.
Jacobson, J. Org. Chem. 1991, 56, 2115; g)P. A. Evans, T. A.
Brandt, Tetrahedron Lett. 1996, 37, 1367; h)H. Sakurai, Y.
Imamoto, T. Hirao, Chem. Lett. 2002, 44; i)B. Hauer, J. F.
Bickley, J. Massue, P. C. A. Pena, S. M. Roberts, J. Skidmore,
Can. J. Chem. 2002, 80, 546.
[12] M. Tiffeneau, A. OrØkhoff, Bull. Soc. Chim. Fr. 1920, 27, 789; L.
ˇ
ˇ
ˇ
ˇ
ˇ
ˇ
Fisnerovµ, B. Kakµc, O. Nemecek, A. Simek, Z. J. Vejdelek,
Collect. Czech. Chem. Commun. 1967, 32, 4082.
[13] For reviews on the use of Weinreb amides in synthesis, see: M.
Mentzel, H. M. R. Hoffmann, J. Prakt. Chem. 1997, 339, 517; J.
Singh, N. Satyamurthi, I. S. Aidhen, J. Prakt. Chem. 2000, 342,
340.
[14] T. W. Greene, P. G. M. Wuts, Protective Groups in Organic
Synthesis, 2nd ed., Wiley, New York, 1991, and references
therein.
[15] C23H19FO3, Mr = 362.38, monoclinic, space group P21/n, a =
13.3598(10), b = 7.1470(4), c = 18.9363(17)Š, b = 97.202(10)8,
V= 1793.8(2)Š 3, Z = 4, 1 = 1.342 gcmꢀ3, m = 0.095 mmꢀ1, T=
291 K, crystal dimensions: 0.25 mm ” 0.3 mm ” 0.3 mm, STOE-
IPDS, MoKa radiation (l = 0.71073 Š), qmax = 25.658, 13715
measured, 3372 unique, and 1924 observed reflections with I >
2s(I), Lorentz polarization correction, direct methods and DF
2
synthesis, minimization of ꢀw(Fo ꢀFc2)2, 320 refined parameters,
2
2
(D/s)max = 0.000, R1[Fo >2s(Fo )] = 0.029, wR2 = 0.059 (all data),
w = 1/[s2(Fo ) + (0.0215P)2] where P = (Fo2 + 2Fc2)/3, S = 1.028,
2
D1max/D1min = 0.122/ꢀ0.106 eŠꢀ3. CCDC-604159 contains the
supplementary crystallographic data for this paper. These data
can be obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
[16] M. Braun, W. Kotter, Angew. Chem. 2004, 116, 520; Angew.
Chem. Int. Ed. 2004, 43, 514.
[17] Although an enantioselectively catalyzed route to 1-methyl-1-
arylindane carboxylic acid has been opened (ref. [16]), it is more
readily prepared from the corresponding esters of (R)- and (S)-
1,1,2-triphenyl-1,2-ethanediol; see: H. Sacha, D. Waldmüller, M.
Braun, Chem. Ber. 1994, 127, 1959.
[18] B. Janowski, S. Wꢁllner, M. Schutkowski, G. Fischer, Anal.
Biochem. 1997, 252, 299.
7458
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7454 –7458