Novel β-dicarbonyl derivatives as inhibitors of aminopeptidase N (APN)

Article abstract of DOI:10.1016/j.bmcl.2013.06.058

Most zinc metalloproteases are over-expressed in tumor cells and play a critical role in the genesis, development, and metastasis of tumors. Novel zinc binding groups (ZBGs) represent a novel strategy to obtain optimal potency and selectivity for zinc metalloproteases inhibitors. Here we described the design, synthesis, and biological studies of novel β-dicarbonyl derivatives as aminopeptidase N (APN/CD13) inhibitors. The results demonstrated that some compounds exhibited moderate to good inhibitory activities against APN with compound 5c being the most potent, suggesting that 5c could serve as new lead for the future APN inhibitor development. The results further confirm our design rationale of β-dicarbonyl moiety as a new ZBG, which may provide a new direction for the design and discovery of zinc metalloproteases inhibitors as new anti-tumor agents.

Full text of DOI:10.1016/j.bmcl.2013.06.058

Bioorganic & Medicinal Chemistry Letters 23 (2013) 4948–4952
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Novel b-dicarbonyl derivatives as inhibitors of aminopeptidase
N (APN)
⇑
Chunhua Ma, Xiaoguang Li, Xuewu liang, Kang Jin, Jiangying Cao, Wenfang Xu
Department of Medicinal Chemistry, School of Pharmaceutical Sciences, Shandong University, Jinan 250012, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 28 February 2013
Revised 5 June 2013
Accepted 20 June 2013
Available online 28 June 2013
Most zinc metalloproteases are over-expressed in tumor cells and play a critical role in the genesis, devel-
opment, and metastasis of tumors. Novel zinc binding groups (ZBGs) represent a novel strategy to obtain
optimal potency and selectivity for zinc metalloproteases inhibitors. Here we described the design, syn-
thesis, and biological studies of novel b-dicarbonyl derivatives as aminopeptidase N (APN/CD13) inhibi-
tors. The results demonstrated that some compounds exhibited moderate to good inhibitory activities
against APN with compound 5c being the most potent, suggesting that 5c could serve as new lead for
the future APN inhibitor development. The results further confirm our design rationale of b-dicarbonyl
moiety as a new ZBG, which may provide a new direction for the design and discovery of zinc metallo-
proteases inhibitors as new anti-tumor agents.
Keywords:
Aminopeptidase N
Zinc binding group
b-Dicarbonyl
Inhibitor
Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved.
Anti-tumor
Most of zinc metalloproteases are significantly important bio-
logical and medical markers for the development of cancer. For
example, matrix metalloproteinases (MMPs), a family of calcium-
and zinc-dependent endopeptidases, have been proved to play cru-
cial roles in tumor proliferation, invasion, angiogenesis, and metas-
tasis;1 Zn2+-dependent HDACs, especially class I and class II
isozymes are associated with tumor proliferation, angiogenesis,
migration, resistance to chemotherapy and preventing apoptosis
phenylbutyric acids (AHPA) scaffold. However, hydrazinocurcu-
min, a synthetic analog of Curcumin with the central b-dicarbonyl
moiety substituted by pyrazole, did not present the enzyme inhib-
itory activity,13 we envisioned that the b-dicarbonyl scaffold may
be able to chelate with the zinc ion which is the crucial catalytic
factor in active site and could potentially serve as a starting point
for the development of novel APN inhibitors with anti-tumor activ-
ity. To examine this hypothesis, a series of novel b-dicarbonyl
derivatives were synthesized, (Fig. 1) the activity results and anal-
ysis of structure activity relationship (SAR) were also shown in this
letter.
Syntheses of target compounds were outlined in Scheme 1.
Optically pure Boc-D-Phe was condensed with Meldrum’s acid to
obtain the key intermediate compound 2. Alcoholysis or aminoly-
sis of 2 with corresponding alcohols or amino acid methyl esters
yielded b-dicarbonyl derivatives 3a–e.14 Removal the Boc protect-
ing group of compounds 3a–d produced 4a–e. Additionally, the
methyl esters of 3c–e were hydrolyzed and then the Boc groups
were removed to yield corresponding acids 5a, 5c–e.15
The target compounds were evaluated for their inhibitory activ-
ities toward APN/CD13 and HDACs as previously described.16
HDACs are zinc metalloproteinase as well as APN and associated
closely with the invasion and metastasis of tumors. The difference
between them is that HDACs mainly exist in nucleus while APN is
an exopeptidase. In order to identify the selectivity of the target
compounds against the two enzymes, the target compounds were
assayed for the inhibitory activities against APN and HDACs, all the
inhibition results were summarized in Tables 1 and 2.
and differentiation;2 the aminopeptidase
N (APN/CD13, EC
3.4.11.2) has been identified as a target for inhibition of tumor vas-
cularization and growth.3,4 Accordingly, it is meaningful to develop
zinc metalloproteases targeted agents for tumor imaging and ther-
apy. One zinc binding group (ZBG) which can coordinate with zinc
ion in the active site of enzyme is indispensable to a successful zinc
metalloproteases inhibitor. Therefore, novel ZBGs are urgently
needed to obtain optimal potency and selectivity zinc metallopro-
teases inhibitors.
APN explored in depth by our laboratory is used as the target to
search for new ZBG. It is a widely expressed type II membrane-
bound metalloprotease that plays a key role in tumor angiogenesis,
growth, and metastasis. Since 1976, several natural inhibitors of
APN have been reported, for example, Bestatin,5 Probestin,6
Amastatin,7 Actinonin,8 Phebestin,9 Lapstatin,10 AHPA-Val,11 Leuhi-
stin,12 Curcumin13 and so on. Amongst these inhibitors, Curcu-
min, which IC50 was 10 lM compared with 2.5 lM of Bestatin, is
a special natural APN inhibitor without 3-Amino-2-hydroxy-4-
⇑
Corresponding author.
E-mail address: wenfxu@gmail.com (W. Xu).
0960-894X/$ - see front matter Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.06.058

C. Ma et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4948–4952
4949
O
O
N
NH
O
O
O
O
HO
OH
HO
OH
Hydrazinocurcumin
Curcumin
β- dicarbonyl may act as ZBG
2
O
O
O
O
R
1
R
O
3
O
N
H
R
NH
NH
O
2
2
1
2
R
= CH , CH Ph
R = side chains of Leu,
3
2
Phe, Phenylglycine
3
R = CH , H
3
Figure 1. The new APNIs to examine the hypothesis that b-dicarbonyl moiety may act as ZBG.
O
OH
O
O
a
OH
Boc
O
HN
1
NH
Boc
O
2
O
O
O
1
2
O
O
O
R
HCl
O
1
R
O
NH
2
c
NH
4a-b
Boc
Boc
b
3a-b
O
O
R
HCl
2
R
N
H
COOCH
3
N
H
COOCH
3
NH
2
NH
3c-e
4c-e
d,c
2
O
O
R
1
2
R
= CH , CH CH
3 5 2
HCl
N
COOH
H
R = CH CH(CH ) ,
NH
2
2
3 2
CH CH , CH
5
2
5
5c-5e
Scheme 1. Reagents and conditions: (a) Meldrum’s acid, EDCIꢀHCl, DMAP, CH2Cl2, 2 h; (b) corresponding alcohols or amino acid methyl esters, acetonitrile, 80 °C; (c) 3N
HCl–EtOAC; (d) NaOH, MeOH, 1 h.

4950
C. Ma et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4948–4952
Table 1
4b, we can find that the activity of benzyl ester is better than that
of methyl ester, which may be due to the system of the aromatic
ring enhancing the interaction with the S10 hydrophobic region
of the APN. Based on the above result, we introduced amino acids
to replace the benzyl ester of 4b in order to increase the interaction
with the hydrophobic region, the activities of these compounds
(4c–e and 5c–e) confirming our design. Comparing 4c–e with
5c–e, we could find that the carboxylic acid derivatives present
better activities than those corresponding methyl ester com-
pounds. This may be due to the fact that the carboxylic acid can in-
crease the water solubility of compounds and interact with S20
pocket by hydrogen bonding. Comparing compounds 5c–e, of
which the R2 groups were different amino acid residues, the APN
inhibitory activities were different owing to the residues. The data
shown in Table 2 suggested that the preferred substitutions against
The structures and inhibitory activities of the target compounds 3a,b and 4a,b against
APN and HDACs
O
O
2
R
O
HN
1
R
a
a
Compd
R1
R2
IC50
(
l
M) APN
IC50 (lM) HDACs
3a
3b
4a
4b
Boc
Boc
H
CH3
CH2Ph
CH3
>1000
>1000
984 310
711 260
5.58 0.55
>1000
>1000
>1000
>1000
>1000
H
CH2Ph
Bestatin
SAHA
Curcumin
0.14 0.13
15.50 4.00
APN were, in decreasing order,
-phenylalanine residues, which was consistent with the result
from ureido derivatives we reported earlier.17
The most active compound is 5c, which has a better inhibitory
activity (IC50 = 1.41 0.25 M) than positive control Bestatin
(IC50 = 5.58 0.55 M) and Curcumin (15.50 4.00 M). Com-
pound 4c, 5d and 5e also dispaly moderate activities (IC50 = 5.04,
4.73, 3.82 M, respectively). The only difference of the structure
L-leucine, L-phenylglycine and
a
Mean values and standard deviations of triplicate experiments are given.
L
l
Table 2
l
l
The structures and inhibitory activities of the target compounds 3c–e, 4c–e and 5c–e
against APN and HDACs
l
O
O
R2
of Beststin and 5c is the hydroxyl group and carbonyl group in
Bestatin replaced by b-dicarbonyl (Fig. 2). While it is interesting
to note that the hydroxyl group and carbonyl group act as the
ZBG of Bestatin, which indicate that the b-dicarbonyl moiety
may interact with APN by chelating the zinc. Therefore, b-dicar-
bonyl moiety can act as ZBG.
To further characterize the affinity of these compounds for Zn
in vitro, a spectrophotometry study was performed.18 In the ab-
sence of zinc acetate, 5c showed absorption maxima at 244.0 nm
which shifted to approximately 266.6 nm when as little as Zn2+
was added. Addition of Zn2+ also increased the peak height by
0.38 OD units. (Fig 3) The results suggested our compounds can
O
N
H
R3
HN
O
R1
a
a
Compd
R1
R2
R3
IC50
(l
M)
IC50 (lM)
APN
HDACs
3c
3d
3e
4c
4d
4e
5c
5d
Boc CH2CH(CH3)2 CH3 >1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
Boc CH2Ph
Boc Ph
CH3 >1000
CH3 >1000
CH2CH(CH3)2 CH3 5.04 2.10
H
H
H
H
H
H
CH2Ph
Ph
CH2CH(CH3)2
CH2Ph
Ph
CH3 7.02 1.30
CH3 6.03 0.48
H
H
H
1.41 0.25
4.73 0.17
3.82 0.45
5.58 0.55
5e
Bestatin
SAHA
Curcumin
0.14 0.13
OH
O
O
H
N
15.50 4.00
N
H
COOH
a
Mean values and standard deviations of triplicate experiments are given.
NH2
O
COOH
NH2
A
B
As shown in Tables 1 and 2, it is worthy to note that most of
b-dicarbonyl derivatives displayed a better enzymatic inhibition
towards APN than that of HDACs, with IC50 values lying in micro-
mole level. These results, to a certain extent, validated our design
strategy for novel potential APNIs. The results may be explained
as follows. In one hand, APN is an exopeptidase which can hydro-
lyze the neutral or basic amino acids from the N-terminal of the
peptide, such as Phe, Tyr, Ala, and Leu. The designed compounds
all have the Phe residue which could be well recognized by APN,
so the compounds may be suitable as APN inhibitors. On the other
hand, the length between hydrophobic group and ZBG of
compounds which is suitable for APN is too short to the slender
channel of HDACs. Thereby the following structure–activity
relationships were mainly focused on APN inhibition.
Among these inhibitors, generally speaking, compounds with
free amino group showed better activity than compounds with
Boc protecting group, suggesting that the free amine is still the suc-
cessful optimization for the inhibition of APN. And it must be
emphasized that the existence of free amine group can endow
the compounds with higher affinity toward the enzyme, which is
consistent with reference and indicates the importance of amino
group in the recognition of inhibitors to APN. Comparing 4a and
Figure 2. The structure of compound Bestatin (A) and 5c (B).
1.2
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
5c
5c+Zn
220 230 240 250 260 270 280 290 300 310
Wavelength (nm)
Figure 3. Absorption spectra of 5c in the presence and absence of zinc.

C. Ma et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4948–4952
4951
Figure 4. (a) The FlexX docking result of 5c with APN (PDB code: 4YFR). (b) The docking result of 5c shown by LIGPLOT.
chelate with Zn2+ in vitro which provided strong evidence for our
assumption.
In summary, one series of novel b-dicarbonyl derivatives as APN
inhibitors had been designed, synthesized and evaluated for the
biological activity. The preliminary results showed that most of
the target compounds represented moderate to good inhibition
and selectivity against APN. The most potent compound, 5c, exhib-
ited a better activity profiles than Bestain which could be served as
new lead for further structure optimization in the future APNIs
research. These results suggest that b-dicarbonyl moiety could
act as ZBG, which provides a new direction for the design and dis-
covery of zinc metalloproteases inhibitors as new antitumor
agents.
Aiming to investigate the interaction between the target com-
pounds and APN, the most active compound 5c, was chosen to
be constructed using a Sybyl/Sketch module and optimized via
Powell’s method by the Tripos force field with convergence
criterion set at 0.05 kcal/(Å mol), and assigned with the Gastei-
ger–Hückel method. The docking study of 5c and the active site
of APN were performed using Sybyl/FlexX module. The active site
was defined as 10.0 Å radius circles around Bestatin in the co-crys-
tal structure (PDB code: 4FYR). Other docking parameters utilized
in the program were remained default. The docking result was
shown in Figure 4a. As diagramed in Figure 4a, the b-dicarbonyl
group can chelate with the zinc ion which is the crucial catalytic
factor in active site. The phenyl group of Phe moiety can insert into
S1 pocket as the same with Bestatin; the hydrophobic residues of
Acknowledgments
This work was supported by National Science and Technology
Specific Projects of China (No. 2011ZX09401-015) and National
Nature Science Foundation of China (No. 21172134).
L
-Leu can plunge into the S10 pocket of APN as the same with Best-
atin. In addition, a 2D pattern of detailed binding mode was
created as well (Fig 4b). The phenyl group of compound 5c can
form hydrophobic interaction with Phe 472 of S1 pocket, while,
the free amine group and carboxylic acid group can form hydrogen
bonds with Gln 213, Glu 355 and Gly 352, Arg 381, respectively.
References and notes
1. Chambers, A. F.; Matrisian, L. M. J. Natl. Cancer Inst. 1997, 89, 1260.
2. Witt, O.; Deubzer, H. E.; Milde, T.; Oehme, I. Cancer Lett. 2009, 277, 8.

4952
C. Ma et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4948–4952
3. Pasqualini, R.; Koivunen, E.; Kain, R.; Lahdenranta, J.; Sakamoto, M.; Stryhn, A.;
Ashmun, R. A.; Shapiro, L. H.; Arap, W.; Ruoslahti, E. Cancer Res. 2000, 60, 722.
4. Arap, W.; Pasqualini, R.; Ruoslahti, E. Science 1998, 279, 377.
5. Umezawa, H.; Aoyagi, T.; Suda, H.; Hamada, M.; Takeuchi, T. J. Antibiot. 1976,
29, 97.
into 1 M hydrochloric acid and extracted with diethyl ether three times. The
organic layer was washed with brine, dried over anhydrous sodium sulfate,
and evaporated to give crude 2. This crude and corresponding alcohols or
amino acid methyl esters were dissolved in acetonitrile (50 ml). After being
heated at 80 °C for 3 h, this mixture was evaporated. The residue was
chromatographed over silica gel to give compound 3. To the stirring solution
of 3 (0.5 g) in 10 ml anhydrous ethyl acetate was added dropwise 5 ml 3 N
EtOAc–HCl and after 2, 3 h, compounds 4 were obtained in the form of hydro-
choloride salt. Filter quickly and dry the cake in vaccum to obtain dried white
solid with high yield. Compound 3 (0.5 g) and 1 N NaOH (4 ml) were add to
4 ml methanol, after 1 h the solution was acidified with 1 N HCl to pH 2,
extracted with EA (30 ml ꢁ3), and subsequently the solvent was removed
under reduced pressure. Then the boc group was removed as above to obtain
compound 5. Compound 5c: yield 30.7%. Mp 120.8–122.2 °C; ESI-MS m/z
[M+1]+ 321.4; 1H NMR (300 MHz, DMSO-d6): d 8.59–8.55 (m, 5H), 7.35-7.23
(m, 5H), 4.49 (m, 1H), 4,25 (m, 1H), 3.85–3.77 (m, 2H), 3.11–2.98 (m, 2H), 1.54
(m, 1H), 1.35 (m, 2H), 0.86 (m, 6H).
6. Aoyagi, T.; Yoshida, S.; Nakamura, Y.; Shigihara, Y.; Hamada, M.; Takeuchi, T. J.
Antibiot. 1990, 43, 143.
7. Rich, D. H.; Moon, B. J.; Harbeson, S. J. Med. Chem. 1984, 27, 417.
8. Umezawa, H. A.; Aoyagi, T.; Tanaka, T.; Suda, H. J. Antibiot. 1985, 38, 1629.
9. Nagai, M.; Kojima, F.; Naganawa, H. J. Antibiot. 1997, 50, 82.
10. Lampret, B. R.; Kidric, J.; Kralj, B. Arch. Microbiol. 1999, 171, 397.
11. Chung, M. C.; Lee, H. J.; Chun, H. K.; Lee, C. H.; Kim, S. I.; Kho, H. Y. Biosci.
Biotechnol. Biochem. 1996, 60, 898.
12. Aoyagi, T.; Yoshida, S.; Matsuda, N.; Ikeda, T.; Hamada, M.; Takeuchi, T. J.
Antibiot. 1991, 44, 573.
13. Shim, J. S.; Kim, J. H.; Cho, H. Y.; Yum, Y. N.; Kim, S. H.; Park, H. J.; Shim, B. S.;
Choi, S. H.; Kwon, H. J. Chem. Biol. 2003, 10, 695.
14. Kikuchi, H.; Sasaki, K.; Sekiya, J., i.; Maeda, Y.; Amagai, A.; Kubohara, Y.;
Oshima, Y. Bioorg. Med. Chem. 2004, 12, 3203.
15. General procedure for the preparation target compounds: To a solution of 1
(2.65 g) in CH2Cl2 (50 ml) was added 1-(3-dimethylaminopropyl)-3-
ethylcarbodiimide hydrochloride (EDCIꢀHCl) (2.88 g), Meldrum’s acid (2.8 g),
and DMAP (1.2 g) at 0 °C. After being stirred for 2 h, the mixture was poured
16. Ma, C.; Jin, K.; Cao, J.; Zhang, L.; Li, X.; Xu, W. Bioorg. Med. Chem. 2013, 21, 1621.
17. Su, L.; Cao, J.; Jia, Y.; Zhang, X.; Fang, H.; Xu, W. ACS Med. Chem. Lett. 2012, 3,
959.
18. Kenneth, H., Urist, I. M. R. In Proceedings of the Society for Experimental Biology
and Medicine. 1962, 109, 797.