Peptidyl hydroxamic acids as methionine aminopeptidase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2003.10.031

A new class of methionine aminopeptidase (MetAP) inhibitors, which contain an internal hydroxamate (N-acyl-N-alkylhydroxylamine) core as the metal-chelating group, has been designed, synthesized, and tested. The compounds exhibited reversible, competitive inhibition against Escherichia coli MetAP as well as human MetAP-1 and MetAP-2. The most potent inhibitor had a K_i value of 2.5 μM and >20-fold selectivity toward E. coli MAP.

Full text of DOI:10.1016/j.bmcl.2003.10.031

Bioorganic & Medicinal Chemistry Letters 14 (2004) 77–79
Peptidyl hydroxamic acids as methionine
aminopeptidase inhibitors
Xubo Hu, Jinge Zhu, Sumant Srivathsan and Dehua Pei*
Department of Chemistry and Ohio State Biochemistry Program, The Ohio State University, 100 West 18th Avenue,
Columbus, OH 43210, USA
Received 28 May 2003; accepted 5 October 2003
Abstract—A new class of methionine aminopeptidase (MetAP) inhibitors, which contain an internal hydroxamate (N-acyl-N-
alkylhydroxylamine) core as the metal-chelating group, has been designed, synthesized, and tested. The compounds exhibited
reversible, competitive inhibition against Escherichia coli MetAP as well as human MetAP-1 and MetAP-2. The most potent inhi-
bitor had a Ki value of 2.5 mM and >20-fold selectivity toward E. coli MAP.
# 2003Elsevier Ltd. All rights reserved.
Protein synthesis universally starts with a methionine
or, in the case of bacteria and organelles, an N-for-
mylmethionine. Following translational initiation, the
N-terminal methionine is frequently removed from nas-
cent polypeptides by methionine aminopeptidase
(MetAP).1 This N-terminal modification has a number
of biological functions including the regulation of pro-
tein stability through the N-end rule pathway and
exposure of the penultimate residue for catalytic func-
tions or post-translational modification. MetAP is an
essential enzyme in all organisms from bacteria to
man.2ꢀ4 Bacteria typically have a single MetAP enzyme,
whereas eukaryotes contain two different MetAP
enzymes, MetAP-1 and MetAP-2.2,5 Recently, it has
been shown that the antiangiogenic activity of fuma-
gillin and ovalicin is due to its inhibition of the pepti-
dase activity of MetAP-2.6,7 This finding has led to
intense efforts in developing MetAP-2 inhibitors as
novel antitumor drugs. There is also a revived interest in
developing MetAP inhibitors as new antibacterial
agents.
Recent experiments with yeast MetAP-1 have suggested
that the in vivo metal may be Zn2+ 14 whereas Mn2+
,
has been proposed as the native metal for human
MetAP-2.15 Reconstituted MetAP containing either
Co2+, Zn2+, or Fe2+ is catalytically active in vitro.11ꢀ16
Since the metal ions are an integral part of the catalytic
machinery, the design of metal chelators has been a
popular approach to developing MetAP inhibitors. The
first class of MetAP inhibitors contain an a-hydroxy-b-
aminoacyl (bestatin) group as the metal ligand.15,17,18
Some of these inhibitors show excellent potency against
human MetAP-2, but are generally of only weak to
moderate activity (Ki in the mM range) against bacterial
MetAP (MetAP-1). A second class contains a triazole
group as the metal ligand.15,19 Interestingly, the potency
of a triazole inhibitor has been shown to be greatly
dependent on the nature of the metal cofactor.15 In this
work, we report a third class of MetAP inhibitors,
which utilize an internal hydroxamic acid to chelate the
metal ions.
The crystal structure of Escherichia coli MetAP bound
with a bestatin inhibitor reveals that the bestatin moiety
is engaged in four interactions with the metal center: (i)
ligation of the N-terminal nitrogen to Co2, (ii, iii) brid-
ging coordination of the a-hydroxyl group to both
metals, and (iv) terminal ligation of Co1 by the keto
oxygen of the pseudo-peptide linkage (Scheme 1, struc-
ture I).17 We envisioned that an N-(a-aminoacyl)hy-
droxylamine should be capable of engaging in similar
interactions with the two-metal center (Scheme 1,
MetAPs are structurally similar metalloenzymes that
contain two divalent metal ions per catalytic unit.8ꢀ10
The nature of the metal ion under physiological condi-
tions is still under debate. Historically, both MetAPs
were thought to be Co2+-dependent enzymes.5,11ꢀ13
* Corresponding author. Tel.: +1-614-688-4068; fax: +1-614-292-
1532; e-mail: pei@chemistry.ohio-state.edu
0960-894X/$ - see front matter # 2003Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.10.031

78
X. Huet al. / Bioorg. Med. Chem. Lett. 14 (2004) 77–79
(N-Benzyloxy)aminopropioic acid (4b) was prepared
from t-butyl acrylic acid, which was first treated with O-
benzylhydroxylamine followed by cleavage of the t-
butyl ester by TFA. Condensation of acids 4a and 4b
with 2-(2-aminoethyl)pyridine gave benzyloxyamines 5a
and 5b, respectively. Amines 5a and 5b were coupled to
Boc-methionine or Boc-norleucine to give peptides 6a–c.
Removal of the protecting groups were effected by cat-
alytic hydrogenation followed by TFA treatment to give
the hydroxamates 1b–d. Compound 1e was prepared in
a fashion similar to 1a.
Scheme 1. Mode of binding by bestatin (I) and hydroxamic acid
inhibitors (II) to MetAP.
structure II). Moreover, the lower pKa of the hydro-
xamic acid relative to that of the bestatin hydroxyl
group may improve the binding affinity of the former to
the MetAP active site. There are numerous examples of
hydroxamate-based metalloenzyme inhibitors. To test
the above notion, we initially designed hydroxamate 1a
(Scheme 2), which contains a methionine as the P1 resi-
due, because MetAP is highly specific for methionine at
this position.20 An N-hydroxyl glycine is placed at the P10
site because MetAP requires a small residue at the penul-
timate position for N-terminal methionyl removal.21
Addition of the 2-(2-pyridyl)ethyl group to the C-ter-
minus of 1a rendered compound 1b a substantially
improved inhibitor. It acted as a competitive inhibitor,
with a Ki value of 2.5 mM against EcMetAP (Fig. 1 and
Table 1). The increased binding affinity is probably due
to both removal of the C-terminal negative charge and
additional interactions brought about by the pyrid-
ylethyl group. On the other hand, replacement of the
P1 methionyl side chain with that of norleucine (n-
butyl) decreased the inhibitor potency by >20-fold
Synthesis of compound 1a is shown in Scheme 2. tert-
Butyl bromoacetate was incubated with O-benzylhy-
droxylamine to give ester 2. Condensation of 2 with N-
Boc-methionine produced amide 3, which was deprot-
ected by catalytic hydrogenation followed by treatment
with trifluoroacetic acid (TFA) to afford the hydrox-
amate 1a. Compound 1a was tested against Co(II)-sub-
stituted E. coli MetAP (EcMetAP) using a coupled
continuous assay.22 Compound 1a behaved as a com-
petitive inhibitor of modest potency, with an inhibition
constant (Ki) of 170 mM (Table 1).
We attributed the poor activity of 1a to the presence of
a negative charge at its C-terminus. In that regard,
MetAP requires at least a tripeptide for efficient
methionyl cleavage.11,12 To remove the negative charge
in 1a, we designed compound 1b, in which a 2-(2-pyr-
idinyl)ethyl group is appended to the C-terminus of 1a.
In addition, we designed compound 1c to assess the
importance of the methionyl side chain to the binding
affinity, and 1d to assess the importance of the linker
length (P10 residue). Compound 1e, which was origin-
ally designed for inhibiting another metalloenzyme, was
used as a control. Compounds 1b–e were synthesized in
a similar fashion as 1a (Scheme 3). N-Benzyloxyglycine
(4a) was obtained by treating compound 2 with TFA. 3-
Scheme 3. Reagents and conditions: (a) For 4a, BrCH2CO2tBu, Et3N,
CH3OH, 60%; for 4b, CH2¼CHCO2tBu, Et3N, EtOH, 70%; (b) TFA-
CH2Cl2, quantitative; (c) 2-(2-aminoethyl)pyridine, HBTU, Et3N,
67% (5a) or 54% (5b); (d) Boc-l-Met-OH or Boc-l-Nle-OH,
ClCO2tBu, Et3N, 62–80%; (e) H2, Pd/C; (f) TFA–CH2Cl2, quantita-
tive yield.
Table 1. Inhibition constants against various Co(II)–MetAPs
Compds
Ki, mMa
EcMetAP
HsMetAP-1
HsMetAP-2
1a
1b
1c
1d
1e
170Æ16
2.5Æ1.0
59Æ6
390Æ32
48Æ5
3 Æ6
ND
1300Æ46
91Æ4
3
164Æ9
ND
48Æ10
8.8Æ2.7
180Æ6
600Æ51
Scheme 2. Reagents and conditions: (a) BnONH2 (0.8 equiv), Et3N
(2 equiv), CH3OH, 60%; (b) Boc-Met-OH/HBTU/Et3N (2 equiv),
71%; (c) H2, Pd/C; (d) TFA.
a Values are meansÆSD from three independent sets of experiments.
ND, not determined.

X. Huet al. / Bioorg. Med. Chem. Lett. 14 (2004) 77–79
79
enzymes, respectively. This work was supported by
grants from the National Institutes of Health (AI40575
to D.P. and GM56495 to Richard C. Holz).
References and notes
1. (a) Meinnel, T.; Mechulam, Y.; Blanquet, S. Biochimie
1993, 75, 1061. (b) Lowther, W. T.; Matthews, B. W.
Biochem. Biophys. Acta 2000, 1477, 157.
2. Li, X.; Chang, Y.-H. Proc. Natl. Acad. Sci. U.S.A. 1995,
92, 12357.
3. Chang, S.-Y. P.; McGary, E. C.; Chang, S. J. Bacteriol.
1989, 171, 4071.
4. Miller, C. G.; Kukral, A. M.; Miller, J. L.; Movva, N. R.
J. Bacteriol. 1989, 171, 5215.
5. Arfin, S. T.; Kendall, R. L.; Hall, L.; Weaver, L. H.;
Stewart, A. T.; Matthews, B. W.; Bradshaw, R. A. Proc.
Natl. Acad. Sci. U.S.A. 1995, 92, 7714.
6. Griffith, E. C.; Su, Z.; Turk, B. E.; Chen, S.; Chang, Y.-
H.; Wu, Z.; Biemann, K.; Liu, J. O. Chem. Biol. 1997, 4,
461.
Figure 1. Lineweaver–Burk plot for the inhibition of EcMetAP by
compound 1b. The lines were fitted to the data according to the
Michaelis–Menten equation: 1=V ¼ 1=Vmax þ KM=ðVmax ½SŠÞ.
ꢁ
(Ki=59 mM for 1c). It was previously reported that
peptides containing N-terminal norleucine are sub-
stantially poorer substrates than the corresponding
methionyl peptides.20 Addition of a methylene group to
the P10 residue had little effect on the overall potency
(Table 1, compare 1c vs 1d). Surprisingly, the control
compound 1e, which contains a d-alanine as the P1
residue, is a significantly better inhibitor than 1a
(Ki=8.8 mM for 1e). Note that the only reported EcMe-
tAP inhibitor, (3R)-amino-(2S)-hydroxyheptanoyl-l-
Ala-l-Val-l-Phe-OMe, contains an n-butyl side chain in
the d-configuration at the P1 position.17
7. Sin, N.; Meng, L.; Wang, M. Q. W.; Wen, J.; Bornmann,
W. G.; Crews, C. M. Proc. Natl. Acad. Sci. U.S.A. 1997,
94, 6099.
8. Roderick, S. L.; Matthews, B. W. Biochemistry 1993, 32,
3907.
9. Tahirov, T. H.; Oki, H.; Tsukihara, T.; Ogasahara, K.;
Yutani, K.; Ogata, K.; Izu, Y.; Tsunasawa, S.; Kato, I. J.
Mol. Biol. 1998, 284, 101.
10. Liu, S.; Widom, J.; Kemp, C. W.; Crews, C. M.; Clardy,
J. Science 1998, 282, 1324.
11. Ben-Basset, A.; Bauer, K.; Chang, S.-Y.; Myambo, K.;
Boosman, A.; Chang, S. J. Bacteriol. 1987, 169, 751.
12. Chang, Y.-H.; Teichert, U.; Smith, J. A. J. Biol. Chem.
1990, 265, 19892.
The above compounds were next tested against the two
human Co(II)-substituted MetAPs, HsMetAP-1 and
HsMetAP-2. In general, these compounds are weaker
inhibitors against the human enzymes (Table 1). In
particular, compounds 1b and 1e exhibited >20-fold
selectivity toward the bacterial enzyme. These results
indicate that it is possible to design inhibitors that are
selective against bacterial MetAP as novel antibiotics.
13. Kendall, R. L.; Bradshaw, R. A. J. Biol. Chem. 1992, 267,
20667.
14. Walker, K. W.; Bradshaw, R. A. Protein Sci. 1996, 7,
2684.
15. Wang, J.; Sheppard, G. S.; Lou, P.; Kawai, M.; Park, C.;
Egan, D. A.; Schneider, A.; Bouska, J.; Lesniewski, R.;
Henkin, J. Biochemistry 2003, 42, 5035.
16. D’souza, V. M.; Holz, R. C. Biochemistry 1999, 38, 11079.
17. Lowther, W. T.; Orville, A. M.; Madden, D. T.; Lim, S.;
Rich, D. H.; Matthews, B. W. Biochemistry 1999, 38,
7678.
18. Mario, J. P., Jr.; Ryan, M. D.; Thompson, S. K.; Veber,
D. F. PCT Int. Appl. 2001, WO0136404; Chem. Abstr.
2001, 134, 764, 367198e.
19. Mario, J. P., Jr.; Thompson, S. K.; Veber, D. F. PCT Int.
Appl. 2001, WO 0124796; Chem. Abstr. 2001, 134, 761,
295827j.
In conclusion, we have shown that N-(a-aminoacyl)hy-
droxylamine derivatives act as metal-chelating inhibi-
tors of both bacterial and humnan MetAPs, with good
to moderate potency. A detailed SAR study of the
inhibitor structures should lead to a new class of potent,
selective MetAP inhibitors as novel antibiotics and
antitumor drugs.
20. Yang, G.; Kirkpatrick, R. B.; Ho, T.; Zhang, G.-F.;
Liang, P.-H.; Johanson, K. O.; Casper, D. J.; Doyle,
M. L.; Mario, J. P., Jr.; Thompson, S. K.; Chen, W.; Tew,
D. G.; Meek, T. D. Biochemistry 2001, 40, 10645.
21. Hirel, P.-H.; Schmitter, J.-M.; Dessen, P.; Fayat, G.;
Blanquet, S. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 8247.
22. Zhou, Y.; Guo, X.-C.; Yi, T.; Yoshimoto, T.; Pei, D.
Anal. Biochem. 2000, 280, 159.
Acknowledgements
We thank Professor Jun O. Liu (Johns Hopkins Uni-
versity) and Professor Richard C. Holz (Utah State
University) for providing human and E. coli MetAP