Structure-based design of a macrocyclic inhibitor for peptide deformylase

Article abstract of DOI:10.1021/jm034113f

A macrocyclic, peptidomimetic inhibitor of peptide deformylase was designed by covalently cross-linking the P1′ and P3′ side chains. The macrocycle, which contains an N-formylhydroxylamine side chain as the metal-chelating group, was synthesized from a diene precursor via olefin metathesis using Grubbs's catalyst. The cyclic inhibitor showed potent inhibitory activity toward Escherichia coli deformylase (Ki = 0.67 nM) and antibacterial activity against both Gram-positive and Gram-negative bacteria (MIC = 0.7-12 μg/mL).

Full text of DOI:10.1021/jm034113f

© Copyright 2003 by the American Chemical Society
Volume 46, Number 18
August 28, 2003
Letters
N-terminal formylation and PDF apparently has no
catalytic function in the mammalian mitochondrion.8
Thus, PDF inhibitors are expected to act as a new class
of broad-spectrum antibacterial agents.
Str u ctu r e-Ba sed Design of a Ma cr ocyclic
In h ibitor for P ep tid e Defor m yla se
Xubo Hu,† Kiet T. Nguyen,†
Christophe L. M. J . Verlinde,‡ Wim G. J . Hol,‡ and
Dehua Pei*,†
PDF is a unique metallopeptidase, which utilizes a
ferrous ion (Fe2+) to catalyze the amide bond hydroly-
sis.9,10 Due to sensitivity of the Fe2+ center to environ-
mental oxygen and other reactive oxygen species, native
PDF is extremely unstable and difficult to work with.11
However, substitution of Ni2+ or Co2+ for the Fe2+ ion
renders the enzyme highly stable while retaining almost
full catalytic activity and substrate specificity of the
native enzyme. Consequently, most of the recent bio-
chemical, structural, and inhibition studies were carried
out with the metal-substituted forms.
Numerous PDF inhibitors have been1 reported in
recent years; essentially all of them are metal chelators.
On the basis of the chelator structure, they can be
classified into three different types: the thiols,12-14 the
hydroxamates,15-19 and the N-formylhydroxylamines or
reverse hydroxamates.20,21 Many of the hydroxamate
and reverse hydroxamate inhibitors exhibit excellent
antibacterial activities in vitro and in animal studies.
One of the reverse hydroxamates from British Biotech
(BB-86398) is currently in phase I clinical trials. How-
ever, since most of these inhibitors still have significant
peptide characteristics, there are some concerns about
their selectivity (e.g., inhibition of matrix metallopro-
teases) and in vivo stability (e.g., proteolysis of the
peptide bonds). A popular approach to improving both
stability against proteolysis and selectivity of peptido-
mimetic inhibitors is to form cyclic peptides or dep-
sipeptides.22,23 We report here the design, synthesis, and
preliminary evaluation of the first macrocyclic inhibitor
of PDF.
Department of Chemistry and Ohio State Biochemistry
Program, The Ohio State University, 100 West 18th Avenue,
Columbus, Ohio 43210, and Department of Biochemistry
and Biomolecular Structure Center, University of
Washington, Seattle, Washington 98195
Received May 26, 2003
Abstr a ct: A macrocyclic, peptidomimetic inhibitor of peptide
deformylase was designed by covalently cross-linking the P1′
and P3′ side chains. The macrocycle, which contains an
N-formylhydroxylamine side chain as the metal-chelating
group, was synthesized from a diene precursor via olefin
metathesis using Grubbs’s catalyst. The cyclic inhibitor showed
potent inhibitory activity toward Escherichia coli deformylase
(KI ) 0.67 nM) and antibacterial activity against both Gram-
positive and Gram-negative bacteria (MIC ) 0.7-12 µg/mL).
The emergence of antibiotic-resistant bacteria has
created an urgent demand for new antibacterial agents
with novel mechanisms of action. Peptide deformylase
(PDF), an essential enzyme involved in bacterial protein
biosynthesis and maturation, is one of the few novel
targets that is currently being pursued for antibacterial
drug design.1-3 In bacteria, protein synthesis starts with
an N-formylmethionine, and as a result, all newly
synthesized polypeptides carry a formylated N-termi-
nus.4 PDF catalyzes the subsequent removal of the
formyl group from the majority of those polypeptides,
many of which undergo further N-terminal processing
by methionine aminopeptidase to produce mature pro-
teins. As an essential activity for survival,5-7 PDF is
present in all eubacteria. On the other hand, protein
synthesis in the eukaryotic cytoplasm does not involve
Structural studies of several PDF-inhibitor com-
plexes20,24,25 have revealed that the inhibitors are bound
in an extended conformation, and the P1′ and P3′ side
chains are similarly oriented. While the P2′ side chain
is extended toward solvent, the P1′ and P3′ side chains
are engaged in intimate interactions with the enzyme.
The P1′ side chain (usually an n-butyl group) fits into a
* To whom correspondence should be addressed. Phone: (614) 688-
4068; fax: (614) 292-1532; e-mail: pei.3@osu.edu.
† The Ohio State University.
‡ University of Washington.
10.1021/jm034113f CCC: $25.00 © 2003 American Chemical Society
Published on Web 08/05/2003

3772 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 18
Letters
Sch em e 1
Sch em e 2a
a
Conditions: (a) pivaloyl chloride, TEA, THF; (b) LiCl, 95% (two
steps); (c) TiCl4, DIPEA, CH2Cl2, (HCHO)n, 52%; (d) H2O2, LiOH,
THF/H2O; (e) BnONH2, HBTU, TEA, CH3CN, 87% (two steps); (f)
DIPAD, Ph3P, THF, 88%; (g) LiOH, i-PrOH/H2O, 93%; (h) HCO2H,
Ac2O, CH2Cl2, quantitative.
deep hydrophobic pocket in the PDF active site. The P3′
side chain makes hydrophobic contacts with a shallow
pocket near the active site as well as one face of the P1′
side chain. It appears that covalent cross-linking of the
P1′ and P3′ side chains would be accommodated by the
PDF active site. Moreover, the rigidity introduced by
cyclization may lock the inhibitor into the PDF-binding
conformation and thus improve binding affinity as well
as selectivity by preventing binding to other enzymes.
On the basis of the structure of reverse hydroxamate 1
(BB-3497) (Scheme 1),20 a potent PDF inhibitor and
antibacterial agent, we designed cyclic compound 2, in
which a nonyl group serves as the cross-linked P1′ and
P3′ side chains (Scheme 1). Molecular modeling indi-
cated that the nonyl group should have the sufficient
length to link the P1′ CR carbon and the P3′ amino
group, while maintaining the extended conformation of
the peptide backbone. We chose tert-leucine as the P2′
residue because other PDF inhibitors containing this
P2′ residue have shown excellent antibacterial activity
and bioavailability.20 Retrosynthetic analysis (Scheme
1) shows that the macrocycle can be conveniently
prepared via olefin metathesis from diene 3, which in
turn can be prepared from acid 4 and amine 5. This
synthetic strategy would allow for the convergent
synthesis of macrocycles of different ring sizes by using
the common acid 4 and varying the length of the
alkenylamino group in 5.
Sch em e 3a
a
Conditions: (a) MsCl, TEA, CH2Cl2; (b) NaN3, DMF-H2O; (c)
LiAlH4, Et2O; 57% (three steps); (d) Boc-tert-Leu-OH, EDC,
CH2Cl2; (e) TFA, 86% (two steps); (f) 4, EDC, CH2Cl2, 89%; (g)
(Pcy3)2Cl2RudCHPh, CH2Cl2, reflux, 83%; (h) Pd/C, H2, MeOH-
EtOAc, quantitative.
Treatment of the resulting amide with TFA resulted in
the amine 5. Coupling of amine 5 with acid 4 gave the
diene 3. The terminal alkenes were cross-linked using
Grubbs’s ruthenium catalyst28,29 to produce the 15-
membered macrocycle 14. The configuration of the ring
CdC bond in this intermediate was not determined.
Catalytic hydrogenation of 14 reduced the double bond
and simultaneously removed the benzyl group from the
N-hydroxyl moiety to give the N-formylhydroxylamine
2. NMR and HPLC analyses indicated a purity of ∼96%.
Compound 2 was assayed against Co(II)-substituted
E. coli PDF using a dehydrogenase assay.30,31 It acted
as a potent inhibitor, with an apparent KI value of 0.67
( 0.2 nM. Thus, cyclization of the P1′ and P3′ side
chains renders compound 2 ∼10-fold more potent than
the acyclic parent compound 1 (IC50 ) 7 nM).20 To gain
Synthesis of acid 4 started from the commercially
available 6-heptenoic acid (Scheme 2). A hydroxymethyl
group was introduced at the C-2 position using 4S-
benzyloxazolidin-2-one as the chiral auxiliary group.26
Removal of the auxiliary group by hydrolysis followed
by condensation with O-benzylhydroxylamine gave the
hydroxamate 9, which was converted into â-lactam 10
through an intramolecular Mitsunobu reaction.27 Hy-
drolysis of the â-lactam furnished acid 11, which was
subsequently formylated at its benzyloxyamine moiety
to give the acid 4.
Synthesis of compound 2 is shown in Scheme 3.
Amine 13, which was obtained from the corresponding
alcohol 12, was condensed with N-Boc-tert-leucine.

Letters
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 18 3773
F igu r e 2. Inhibition of B. subtilis (A) and E. coli (B) cell
growth by inhibitor 2. An overnight culture was diluted 1000-
fold into 2 mL of fresh LB medium containing the specified
concentrations of inhibitor 2 and incubated at 37 °C. Cell
densities were measured at the specified times on a UV-vis
spectrophotometer.
F igu r e 1. (A) Electronic absorption spectra of Co(II)-
substituted PDF (240 µM) in the absence and presence of
inhibitor 2 (500 µM). (B) Model showing the binding mode of
inhibitor 2 to the E. coli PDF active site. Modeling was carried
out by docking compound 2 into the structure of E. coli PDF
bound with reverse hydroxamate inhibitor 1 (PDB access code
1G27). Protein residues involved in hydrophobic interactions
with the inhibitor nonyl ring, as well as the metal ligands,
are labeled (atom color code: protein C ) gray, inhibitor C )
caramel, N ) blue, O ) red, metal ) purple; metal bonds )
green; hydrogen bonds ) purple).
The in vitro antibacterial activity of the cyclic inhibi-
tor was tested against E. coli and Bacillus subtilis, the
representative Gram-negative and Gram-positive bac-
teria, respectively. Figure 2 shows the bacterial cell
growth curves in the presence of varying concentrations
of inhibitor 2. Compound 2 exhibited potent antibacte-
rial activity against B. subtilis, with a minimal inhibi-
tory concentration (MIC) of 2-4 µM (or 0.7-1.4 µg/mL).
It is only moderately active against E. coli, with an MIC
of ∼32 µM (∼12 µg/mL). The lower activity against E.
coli is likely due to its inefficient permeation of the
bacterial outer membrane and/or being removed from
the cells by the efflux pump.
insight into the mechanism of inhibition, the Co-PDF-
inhibitor 2 complex was examined by UV-visible spec-
troscopy. Binding of the inhibitor resulted in marked
blue shift (by ∼40 nm) and reduction in the maximum
intensity of the D-D transition bands in the absorption
spectrum of the cobalt ion (Figure 1a). This result
suggests that compound 2 is directly ligated to the metal
ion. The maximum absorptivity of ∼200 M-1 cm-1 for
the PDF-inhibitor complex is consistent with a biden-
tate interaction between the N-formylhydroxylamine
group and the metal and the formation of a pentacoor-
dinated cobalt.32 Molecular modeling showed that in-
hibitor 2 fits snugly in the active site of E. coli PDF
(Figure 1b). The N-formylhydroxylamine is coordinated
with the metal ion via both oxygen atoms. There are
three hydrogen bonds formed between the protein
and the inhibitor: from Ile-44 backbone amide to
P1′ carbonyl, from P2′ amide to Gly-89 carbonyl, and
from Gly-89 amide to P2′ carbonyl group. The nonyl
group is engaged in extensive hydrophobic interactions
with protein side chains including those of Ile-44,
Ile-86, Glu-88, Ile-128, Cys-129, and His-132. Overall,
these interactions are very similar to those observed in
the PDF-inhibitor 1 complex.20
In conclusion, based on our earlier observations that
the P1′ and P3′ side chains of PDF inhibitors are closely
packed in the PDF-inhibitor complex, we have devel-
oped a new class of macrocyclic PDF inhibitor by
covalently linking the two side chains. The cyclic inhibi-
tor is highly potent against PDF and has excellent to
moderate antibacterial activity against both Gram-
positive and Gram-negative bacteria. This result dem-
onstrates that cyclization of the P1′ and P3′ side chains
is a viable approach to developing potent PDF inhibi-
tors. Due to their more rigid structures, cyclic inhibitors
of this type may also have improved stability and
selectivity. Further evaluation of inhibitor 2 with regard
to its selectivity and in vivo stability as well as SAR
studies of macrocycles of different ring sizes are already
underway in our laboratories. These cyclic inhibitors
may provide a novel class of antibiotics.

3774 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 18
Letters
deformylase inhibitors and antibacterial agents. J . Med. Chem.
2000, 43, 2324-2331.
(17) J ayasekera, M. M. K.; Kendall, A.; Shammas, R.; Dermyer, M.;
Tomala, M.; Shapiro, M. A.; Holler, T. P. Novel nonpeptidic
inhibitors of peptide deformylase. Arch. Biochem. Biophys. 2000,
381, 313-316.
(18) Thorarensen, A.; Douglas, M. R., J r.; Rohrer, D. C. et al.
Identification of novel potent hydroxamic acid inhibitors of
peptidyl deformylase and the importance of the hydroxamic acid
functionality on inhibition. Bioorg. Med. Chem. Lett. 2001, 11,
1355-1358.
Ack n ow led gm en t. This work was supported by a
grant from the National Institutes of Health (AI40575).
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails for the synthesis and assay of compound 2 and the
spectral data. This material is available free of charge via the
Internet at http://pubs.acs.org.
Refer en ces
(19) Roblin, P. M.; Hammerschlag, M. R. In vitro activity of a new
antibiotic, NVP-PDF386 (VRC4887), against Chlamydia pneu-
moniae. Antimicrob. Agents Chemother. 2003, 47, 1447-1448.
(20) Clements, J . M.; Beckett, R. P.; Brown, A.; Catlin, G.; Lobell,
M.; Palan, S.; Thomas, W.; Whittaker, M.; Wood, S.; Salama,
S.; Baker, P. J .; Rodgers, H. F.; Barynin, V.; Rice, D. W.; Hunter,
M. G. Antibiotic activity and characterization of BB-3497, a novel
peptide deformylase inhibitor. Antimicrob. Agents Chemother.
2001, 45, 563-570.
(21) Smith, H. K.; Beckett, R. P.; Clements, J . M.; Doel, S.; East, S.
P.; Launchbury, S. B.; Pratt, L. M.; Spavold, Z. M.; Thomas, W.;
Todd, R. S.; Whittaker, M. Structure-activity relationship of the
peptide deformylase inhibitor BB-3497: modification o9f the
metal binding group. Bioorg. Med. Chem. Lett. 2002, 12, 3595-
3599.
(1) Giglione, C.; Pierre, M.; Meinnel, T. Peptide deformylase as a
target for new generation, broad spectrum antimicrobial agents.
Mol. Microbiol. 2000, 36, 1197-1205.
(2) Pei, D. Peptide deformylase: a target for novel antibiotics?
Emerging Ther. Targets 2001, 5, 23-40.
(3) Yuan, Z.; Trias, J .; White, R. J . Deformylase as
a novel
antibacterial target. Drug Discovery Today 2001, 6, 954-961.
(4) Meinnel, T.; Mechulam, Y.; Blanquet, S. Methionine as trans-
lational start signal: a review of the enzymes of the pathway in
Escherichia coli. Biochimie 1993, 75, 1061-1075.
(5) Mazel, D.; Pochet, S.; Marliere, P. Genetic characterization of
polypeptide deformylase, a distinctive enzyme of eubacterial
translation. EMBO J . 1994, 13, 914-923.
(6) Meinnel, T.; Blanquet, S. Characterization of the Thermus
thermophilus locus encoding peptide deformylase and methionyl-
tRNA(fMet) formyltransferase. J . Bacteriol. 1994, 176, 7387-
7390.
(7) Margolis, P. S.; Hackbarth, C. J .; Young, D. C.; Wang, W.; Chen,
D.; Yuan, Z.; White, R.; Trias, J . peptide deformylase in S.
aureus: resistance to inhibition is mediated by mutations in the
formyltransferase gene. Antimicrob. Agents Chemother. 2001,
44, 1825-1831.
(8) Nguyen, K. T.; Hu, X.; Colton, C.; Chakratarti, R.; Zhu, M. X.;
Pei, D. Characterization of a human peptide deformylase:
Implications for antibacterial drug design. Biochemistry (in
press).
(9) Rajagopalan, P. T. R.; Yu, X. C.; Pei, D. Peptide deformylase: a
new type of mononuclear iron protein. J . Am. Chem. Soc. 1997,
119, 12418-12419.
(10) Groche, D.; Becker, A.; Schlichting, I.; Kabasch, W.; Schultz, S.;
Wagner, A. F. V. Isolation and crystallization of functionally
competent Escherichia coli peptide deformylase forms containing
either iron or nickel in the active site. Biochem. Biophys. Res.
Commun. 1998, 246, 342-346.
(22) Xue, C.-B.; He, X.; Roderick, J . et al. Design and synthesis of
cyclic inhibitors of matrix metalloproteases and TNF-R produc-
tion. J . Med. Chem. 1998, 41, 1745-1748.
(23) Wei, C.-Q.; Gao, Y.; Lee, K. et al. Macrocyclization in the design
of Grb2 SH2 domain-binding ligands exhibiting high potency in
whole-cell systems. J . Med. Chem. 2003, 46, 244-254.
(24) Hao, B.; Gong, W.; Rajagopalan, P. T. R.; Zhou, Y.; Pei, D.; Chan,
M. K. Structural basis for the design of antibiotics targeting
peptide deformylase. Biochemistry 1999, 38, 4712-4719.
(25) Guilloteau, J .-P.; Mathieu, M.; Giglione, C.; Blanc, V.; Dupuy,
A.; Chevrier, M.; Gil, P.; Famechon, A.; Meinnel, T.; Mikol, V.
The crystal structures of four peptide deformylases bound to the
antibiotic actinonin reveal two distinct types: A platform for the
structure-based design of antibacterial agents. J . Mol. Biol. 2002,
320, 951-962.
(26) Evans, D. A.; Rieger, D. L.; Bilodeau, M. T.; Urpi, F. Stereose-
lective aldol reactions of chlorotitanium enolates. An efficient
method for the assemblage of polypropionate-related synthons.
J . Am. Chem. Soc. 1991, 113, 1047-1049.
(11) Rajagopalan, P. T. R.; Pei, D. Oxygen-mediated inactivation of
peptide deformylase. J . Biol. Chem. 1998, 273, 22305-22310.
(12) Meinnel, T.; Patiny, L.; Ragusa, S.; Blanquet, S. Design and
synthesis of substrate analogue inhibitors of peptide deformy-
lase. Biochemistry 1999, 38, 4287-4295.
(27) Mitsunobu, O. The use of diethyl azodicarboxyalte and triph-
enylphosphane in synthesis and transformation of natural
products. Synthesis 1981, 1-28.
(28) Miller, S. J .; Blackwell, H. E.; Grubbs, R. H. Application of
ringclosing metathesis to the synthesis of rigidified amino acids
and peptides. J . Am. Chem. Soc. 1996, 118, 9606-9614.
(29) Furstner, A. Olefin metathesis and beyond. Angew. Chem., Int.
Ed. 2000, 39, 3012-3043.
(30) Rajagopalan, P. T. R.; Datta, A.; and Pei, D. Purification,
characterization, and inhibition of peptide deformylase from
Escherichia coli. Biochemistry 1997, 36, 13910-13918.
(31) Lazennec, C.; Meinnel, T. Formate dehydrogenase-coupled spec-
trophotometric assay of peptide deformylase. Anal. Biochem.
1997, 244, 180-182.
(32) Bertini, I.; Luchinat, C. The reaction pathways of zinc enzymes
and related biological catalysts. In Bioinorganic Chemistry;
Bertini, I., Gray, H. B., Lippard, S. J ., Valentine, J . S., Eds.;
University Science Books: Sausalito, CA, 1994; pp 37-106.
(13) Huntington, K. M.; Yi, T.; Wei, Y.; Pei, D. Synthesis and
antibacterial activity of peptide deformylase inhibitors. Bio-
chemistry 2000, 39, 4543-4551.
(14) Wei, Y.; Yi, T.; Huntington, K. M.; Chaudhury, C.; Pei, D.
Identification of a potent peptide deformylase inhibitor from a
rationally designed combinatorial library. J . Comb. Chem. 2000,
2, 650-657.
(15) Chen, D. Z.; Patel, D. V.; Hackbarth, C. J .; Wang, W.; Dreyer,
G.; Young, D.; Margolis, P. S.; Wu, C.; Ni, Z.-J .; Trias, J .; White,
R.; Yuan, Z. Actinonin, a naturally occurring antibacterial agent,
is a potent deformylase inhibitor. Biochemistry 2000, 39, 1256-
1262.
(16) Apfel, C.; Banner, D. W.; Bur, D.; Dietz, M.; Hirata, T.;
Hubschwerlen, C.; Locher, H.; Page, M. G. P.; Pirson, W.; Rosse,
G.; Specklin, J .-L. Hydroxamic acid derivatives as potent peptide
J M034113F