Inhibitors of the FEZ-1 metallo-β-lactamase

Article abstract of DOI:10.1016/j.bmcl.2006.11.053

Metallo-β-lactamases (MBLs) catalyze the hydrolysis of β-lactams including penicillins, cephalosporins and carbapenems. Starting from benzohydroxamic acid (1) structure-activity studies led to the identification of selective inhibitors of the FEZ-1 MBL, e.g., 2,5-substituted benzophenone hydroxamic acid 17 has a K_i of 6.1 ± 0.7 μM against the FEZ-1 MBL but does not significantly inhibit the IMP-1, BcII, CphA or L1 MBLs.

Full text of DOI:10.1016/j.bmcl.2006.11.053

Bioorganic & Medicinal Chemistry Letters 17 (2007) 964–968
Inhibitors of the FEZ-1 metallo-b-lactamase
a
b
b
b
`
´
Benoˆıt M. R. Lienard, Louise E. Horsfall, Moreno Galleni, Jean-Marie Frere
and Christopher J. Schofielda,*
aChemistry Research Laboratory, 12 Mansfield Road, Oxford, United Kingdom
` `
´
Centre d’Inge´nie´rie des Prote´ines, Universite´ de Liege, Allee de 6 Aout B6, Sart-Tilman, Liege, Belgium
b
Received 7 October 2006; revised 13 November 2006; accepted 14 November 2006
Available online 18 November 2006
Abstract—Metallo-b-lactamases (MBLs) catalyze the hydrolysis of b-lactams including penicillins, cephalosporins and carbapen-
ems. Starting from benzohydroxamic acid (1) structure-activity studies led to the identification of selective inhibitors of the
FEZ-1 MBL, e.g., 2,5-substituted benzophenone hydroxamic acid 17 has a Ki of 6.1 0.7lM against the FEZ-1 MBL but does
not significantly inhibit the IMP-1, BcII, CphA or L1 MBLs.
Ó 2006 Elsevier Ltd. All rights reserved.
b-Lactamases can be classified into those employing a
nucleophilic serine (classes A, C and D) and those bind-
ing one or two zinc ions (the metallo-b-lactamases,
MBLs, class B) at their active site. To date, medical
attention has focused on the serine active BLs and inhib-
itors of these enzymes are widely used. However, MBLs
hydrolyse most b-lactam antibiotics and represent a
growing clinical threat (for review, see Refs. 1,2).
MBL sub-classes B1 and B3 display a broad substrate
selectivity, whereas sub-class B2 are reported to efficient-
ly hydrolyse only carbapenems.3,4 Sub-class B1 MBLs,
including BcII (Bacillus cereus) and IMP-1 (Pseudomo-
nas aeruginosa), accept b-lactam substrates with one or
two active site Zn(II) ions, whereas B3 MBLs, e.g.,
FEZ-1 (Legionella gormanii) and GOB-1 (Chryseobacte-
rium meningosepticum), are only active with two Zn(II)
ions. The B2 sub-class MBL CphA (Aeromonas hydro-
philia) is only active as a monozinc protein and is non-
competitively inhibited by the presence of a second
Zn(II).5 MBL inhibitors can be classified into those
containing thiols (e.g., thiomandelic acid,6 D-captopril,7
6-(mercaptomethyl)penicillinates8 and homocysteinyl
peptides9) and non-thiol-containing compounds (e.g.,
trifluoromethylketones,10 hydroxamic acids,11 biphenyl
tetrazoles,12 N-arylsulfonyl hydrazones13 and 2,3-disub-
stituted succinic acids14). Notably, only thiomandelate
has been reported to inhibit (non-selectively) FEZ-1
with a Ki value of 0.27lM.6 Despite similarities in its
overall fold, crystallographic analyses indicate that the
active site of the FEZ-1 MBL is significantly different
to that of other sub-class B3 MBLs,15 possibly reflecting
difficulties in obtaining inhibitors for it. To demonstrate
that FEZ-1 inhibitors can be prepared, and with the aim
of understanding differences between FEZ-1 and other
MBLs, efforts were made to identify non-thiol FEZ-1
inhibitors. We report the identification and synthesis
of benzophenone hydroxamic acids as FEZ-1 selective
inhibitors.
Initial studies, based on the screening of a set of small
molecules bearing potential metal binding chelation
functions, led to the observation that benzohydroxamic
acid (1, Fig. 1) inhibited FEZ-1 (55% of inhibition at
100lM under standard assay conditions), but did not
inhibit IMP-1, BcII and CphA. Furthermore, 1 did not
inhibit the L1 MBL, which, like FEZ-1, is from sub-class
B3. The synthesis of derivatives of 1 was then carried out
to identify more potent FEZ-1 inhibitors. Hydrox-
ybenzohydroxamic acids (2a–c) were prepared from the
corresponding hydroxybenzoic acids in five steps (in
overall yields of 17–52%; see supporting information)
and screened for inhibition against IMP-1, BcII, CphA,
L1 and FEZ-1. Compounds 2a–c inhibited FEZ-1 with
ca. 45%, 64% and 53% inhibition observed in the stan-
dard assay. Hydroxamic acids 3 and 4 did not significant-
ly (<10%) inhibit FEZ-1 nor the other tested MBLs,
arguing against the non-specific inhibition of FEZ-1 by
Keywords: Antibiotic; Inhibition; Metallo-b-lactamases; Structure–
activity relationship.
*
Corresponding author. Tel.: +44 1865 275625; fax: +44 1865
275674; e-mail: Christopher.schofield@chem.ox.ac.uk
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.11.053

´
B. M. R. Lienard et al. / Bioorg. Med. Chem. Lett. 17 (2007) 964–968
965
Figure 1. Derivatives of benzohydroxamic acid 1 synthesized as potential MBL inhibitors.
Zn(II) complexation in solution. Solution kinetics with
FEZ-1 revealed competitive inhibition for the most po-
tent inhibitor 2b with a Ki of 17.9 0.7lM.
Since a meta-hydroxyl substituent on the phenyl ring ap-
peared to somewhat enhance the inhibition against
FEZ-1 (from 55% inhibition for 1 to 64% for 2b) whilst
retaining selectivity, dihydroxy-derivatives of 2b were
targeted. Compounds 5a–d (Fig. 1) were synthesized
via a similar route used for 2a–c (see supporting infor-
mation). Despite retaining FEZ-1 selectivity, the pres-
ence of an additional hydroxyl group at the C-2 (5a),
C-4 (5b) or C-6 (5d) positions on the aromatic ring
resulted in an apparent reduction in potency compared
to 2b. However, 3,5-dihydroxylated compound 5c
(Ki = 13.3 1.0lM) had about the same potency as 2b
(17.9 0.7lM) and was FEZ-1-selective. The apparent
slight increase in the potency on derivatization at C-5
prompted the investigation of further 3-hydroxy-5-
substituted benzohydroxamic acids.
5-Amidobenzohydroxamic acids 6a–e and 5-carbo-
xybenzohydroxamic acid 7 were prepared from dimethyl
ester 8, in seven steps and in overall yields of 26–57% for
6a–e and in six steps and 41% yield for 7 (Scheme 1).
Thus, methyl isophthalate 8 was alkylated with allyl
bromide to give ether 9. Mono-saponification of the
methyl isophthalate ester 9 to generate 10 was followed
by EDCI/HOBt-mediated coupling of O-benzylhydr-
oxylamine HCl to afford 11. Hydrolysis of the methyl es-
ter of 11 afforded carboxylic acid 12 which was then
coupled with aromatic, aliphatic and cyclic amines to af-
ford benzamides 13a–e. Deprotection of allyl ethers
13a–e catalyzed by tetrakis(triphenylphosphine)palladi-
um(0) gave phenols 14a–e, which were converted by
palladium-catalyzed-hydrogenolysis to the target
hydroxamic acids 6a–e. Starting from protected ben-
zohydroxamic acid 11, palladium-mediated allyl depro-
tection afforded phenol 15, which was followed by
methyl ester hydrolysis to give 16, and palladium-cata-
lyzed hydrogenolysis afforded 7.
Scheme 1. Reagents and conditions: (a) Allyl bromide, Cs2CO3, DMF,
1 h, 98%; (b) NaOH, MeOH/THF, 1 h, 69%; (c) H2NOBnÆHCl, HOBt,
EDCI, Et3N, CH2Cl2, 12 h, 87%; (d) NaOH, THF, 3 h, 87%; (e)
HNR1R2, HOBt, EDCI, Et3N, THF, 6 h, 68–98%; (f) Pd(PPh3)4,
Pyrrolidine, CH2Cl2, 20 min, 61–90%; (g) 10% Pd-C, H2, MeOH,
1–16 h, 76–99%.
Benzophenone hydroxamic acid 17 was prepared from
dimethyl ester 18, in 10 steps and overall yield of 8%
(Scheme 2). Dimethyl ester 18 was mono-saponified
using KOH in DMSO to give monomethyl ester 19. For-
mation of acid chloride 20 (SOCl2), followed by Friedel–
Crafts acylation, afforded benzophenone 21 (para/ortho
regioselectivity, 96:4). Methyl ester 21 (as an para/ortho
mixture) was hydrolysed to give 22 and then demethylat-
ed using an ionic liquid solution (trimethylamine HCl
and aluminium chloride16) to give hydroxybenzophe-
none 23a.
At this stage, the ortho isomer 23b was isolated. Protec-
tion of 23a using allyl bromide and cesium carbonate in

´
B. M. R. Lienard et al. / Bioorg. Med. Chem. Lett. 17 (2007) 964–968
966
Strikingly, when the 5-hydroxyl group of compound 5c
was replaced by a benzoyl moiety to generate benzophe-
none hydroxamate 17, FEZ-1 was inhibited by 85%
compared with the absence of inhibitor. Moreover, 17
remained selective for FEZ-1, with no significant inhibi-
tion detected against IMP-1, BcII, CphA or even L1.
The Ki value for 17 of 6.1 0.7 lM represents a three
fold increase in potency compared to 2b without reduc-
tion of selectivity. A significant drop of inhibitory activ-
ity was observed for carboxylate 23a with ca. 33% of
inhibition compared to hydroxamate 17 (85%). This re-
sult indicates that the hydroxamic acid function is
important for efficient FEZ-1 inhibition (probably via
interactions with the Zn(II) ion(s)) and that the benzo-
phenone core may interact favourably with FEZ-1.
Scheme 2. Reagents and conditions: (a) KOH, DMSO, 130 °C, 3h,
49%; (b) SOCl2, reflux, 1h, >95%; (c) anisole, AlCl3, CH2Cl2, 3h, 72%
for para isomer, [96:4] para/ortho; (d) NaOH, MeOH, 60 °C, 1h, 99%;
(e) TMAH-Al2Cl7, CH2Cl2, reflux, 17h, 95%; (f) Allyl bromide,
Cs2CO3, DMF, rt, 1h, 96%; (g) NaOH, MeOH, 60 °C, 1h, 95%; (h)
HOBt, EDCI, Et3N, H2NOBn, CH2Cl2, rt, 16h, 67%; (i) Pd(PPh3)4,
Pyrrolidine, CH2Cl2, rt, 1h, 38%; (j) 10% Pd-C, MeOH, H2, rt, 1h,
>95%.
DMF gave triallyl-protected benzophenone 24, and
saponification of the allyl ester afforded carboxylic acid
25. Generation of benzyl-protected hydroxamic acid 26
was accomplished using the EDCI/HOBt-mediated
coupling of O-benzylhydroxylamine HCl in CH2Cl2.
Deprotection, first with tetrakis(triphenylphosphine)pal-
ladium(0) (to give 27) and then 10% palladium(0) on
charcoal and H2 gave 17.
Replacement of the 5-hydroxyl group of 5c with a
carboxylate (7) almost removed activity versus FEZ-1
(Table 1). However, amide derivatives of 7 (6a–e)
retained some potency versus FEZ-1 with 58% (6a),
54% (6b), 36% (6c), 34% (6d) and 53% (6e) inhibition
being observed in the standard assay. These observa-
tions prompted investigation into alternative hydro-
phobic more rigid aromatic derivatives such as in
benzophenones 17 and 23a.
Figure 2. Deconvoluted ESI-MS spectra of (a) the FEZ-1 MBL and a
1/1 mixture of FEZ-1 and compound 17 at a sample cone voltage
varied from (b) 80V to (c) 120V.
Table 1. Residual activity of various members of the MBL family after treatment with compounds 7, 6a–e, 23a and 17 (100lM)
Inhibitor
R1
R2
% Activity of IMP-1
% Activity of BcII
% Activity of CphA
% Activity of L1
100
97 19
% Activity of FEZ-1
100
92 12
None
7
6a
—
—
H
—
—
100
67
4
100
93
2
100
79
1
8
6
9
3
8
4
3
6
9
5
9
Ph
CH2Ph
92
81
82
90 18
103
95
2
42
46
64
66
47
67
15
2
3
6
7
1
7
4
6b
6c
H
103 10
100
81
86
90
92
5
1
1
5
3
3
1
4
Pyrrolidine
Et
85
85
5
92
93
6d
6e
Et
H
101 10
103 10
CH3
—
93
100
96
107
23a
17
—
—
7
86 4
91 11
83 13
84 15
1
—
83 13
102 14

´
B. M. R. Lienard et al. / Bioorg. Med. Chem. Lett. 17 (2007) 964–968
967
Figure 3. (a) Highest GOLD score docking conformation representing the possible binding mode of 17 to FEZ-1 (PDB code 1K07);15 (b) view from
the crystal structure of aminopeptidase(ZnII)2-para-iodo-D-phenylalanine hydroxamate (PDB code 1IGB).19 The similarity of the hydroxamate-
Zn(II) interactions between the docked FEZ-1(ZnII)2-17 and the crystal structure of aminopeptidase(ZnII)2-PIPAH complexes is highlighted. Factors
that may contribute to the proposed binding mode of FEZ-1(ZnII)2-17 include electrostatic interactions between the hydroxamic acid of 17 and the
Zn(II) ions, hydrogen bonding between Tyr156 and C-3 hydroxyl groups of 17, hydrogen bonding between the Lys297 side chain and the C-11
hydroxyl group of 17 and hydrophobic interactions between Pro226 and 17 (see supporting information for additional views).
To investigate the binding of compound 17 to FEZ-1,
soft ionization electrospray mass spectrometry (ESI-
MS) experiments were carried out (Fig. 2a). Peaks at
ca. 29,435 Da (A) and ca. 29,567 Da (B) were observed;
they correspond to FEZ-1(ZnII)2 and FEZ-
1(ZnII)2 + methionine (met-FEZ-1(ZnII)2) complexes,
respectively. The assignment of the N-terminal Met
residue was confirmed by tryptic digests followed by
MALDI-MS analyses.
with more hydrophobic functionalities at C-5 (e.g., 6a–e
and 23a) may be more suited for interaction with
Pro226 leading to an increase in potency.
Comparison of the predicted binding mode for 17 with
FEZ-1 MBL, with crystal structures of wildtype BcII
and CphA (see supporting information), suggests that
the much narrower active sites of BcII and CphA may
hinder binding of the para-hydroxybenzoyl substituent
of compound 17 possibly rationalizing the reason for
the lack of inhibition of 17 against IMP-1, BcII, CphA
and L1.
Enzyme–inhibitor complexes were not observed for the
less potent inhibitors 1, 2b and 5c, but one molecule of
17 was observed to bind to FEZ-1 (Fig. 2b) correspond-
ing to a mass shift of ca. 270 Da relative to the unligand-
ed enzyme; in agreement with the MW of 17 (273 Da).
The complex of 17 with FEZ-1 was not detected upon
increasing the sample cone from 80 V (Fig. 2b) to
120 V (Fig. 2c), indicative of non-covalent binding.
Acknowledgments
The authors thank the EU for support B.M.R.L. and
L.E.H. (Contract No. HPRN-CT-2002-00264) and
Amura for a CASE award to BMRL.
Although there are caveats with respect to conformation-
al changes and metal ligation, docking analyses of 17
using GOLD,17 which has been previously used with
MBLs,18 gave 10 conformations with seven displaying
identical binding modes to FEZ-1 (Fig. 3a). The predicted
interactions between the carbonyl and hydroxyl oxygen
atoms of the hydroxamate of 17 with the FEZ-1 Zn(II)
ions are similar to the crystallographically observed
interactions in the complex of a dizinc-dependent amino-
peptidase (Fig. 3b) and para-iodo-D-phenylalanine
hydroxamate.19 The benzophenone core of 17 is predicted
to adopt a twisted shape around Pro226, apparently max-
imizing the interaction with the benzophenone carbonyl
and the aromatic C-4 of 17. This interaction may explain
the lack of FEZ-1 activity for 3, 4 and 7 since the presence
of a negatively charged carboxylate function at C-5 may
be incompatible with Pro226. In contrast, compounds
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2006.11.053.
References and notes
1. Bush, K.; Jacoby, G. A.; Medeiros, A. A. Antimicrob.
Agents Chemother. 1995, 39, 1211.
2. Phillips, O. A. Expert Opin. Ther. Pat. 2006, 16, 319.
3. Felici, A.; Amicosante, G.; Oratore, A.; Strom, R.;
`
Ledent, P.; Joris, B.; Fanuel, L.; Frere, J. M. Biochem.
J. 1993, 291(Pt 1), 151.

´
B. M. R. Lienard et al. / Bioorg. Med. Chem. Lett. 17 (2007) 964–968
968
4. Segatore, B.; Massidda, O.; Satta, G.; Setacci, D.;
Amicosante, G. Antimicrob. Agents Chemother. 1993,
37, 1324.
5. Hernandez Valladares, M.; Felici, A.; Weber, G.; Adolph,
H. W.; Zeppezauer, M.; Rossolini, G. M.; Amicosante,
Rossolini, G. M.; Schofield, C. J. Bioorg. Chem. 1999, 27,
35.
12. Toney, J. H.; Fitzgerald, P. M.; Grover-Sharma, N.;
Olson, S. H.; May, W. J.; Sundelof, J. G.; Vanderwall, D.
E.; Cleary, K. A.; Grant, S. K.; Wu, J. K.; Kozarich, J.
W.; Pompliano, D. L.; Hammond, G. G. Chem. Biol.
1998, 5, 185.
13. Siemann, S.; Evanoff, D. P.; Marrone, L.; Clarke, A. J.;
Viswanatha, T.; Dmitrienko, G. I. Antimicrob. Agents
Chemother. 2002, 46, 2450.
`
G.; Frere, J. M.; Galleni, M. Biochemistry 1997, 36, 11534.
6. Mollard, C.; Moali, C.; Papamicael, C.; Damblon, C.;
Vessilier, S.; Amicosante, G.; Schofield, C. J.; Galleni, M.;
`
Frere, J. M.; Roberts, G. C. J. Biol. Chem. 2001, 276,
45015.
7. Heinz, U.; Bauer, R.; Wommer, S.; Meyer-Klaucke, W.;
Papamichaels, C.; Bateson, J.; Adolph, H. W. J. Biol.
Chem. 2003, 278, 20659.
8. Buynak, J. D.; Chen, H.; Vogeti, L.; Gadhachanda, V.
R.; Buchanan, C. A.; Palzkill, T.; Shaw, R. W.;
Spencer, J.; Walsh, T. R. Bioorg. Med. Chem. Lett.
2004, 14, 1299.
14. Toney, J. H.; Hammond, G. G.; Fitzgerald, P. M.;
Sharma, N.; Balkovec, J. M.; Rouen, G. P.; Olson, S.
H.; Hammond, M. L.; Greenlee, M. L.; Gao, Y. D.
J. Biol. Chem. 2001, 276, 31913.
15. Garcia-Saez, I.; Mercuri, P. S.; Papamicael, C.; Kahn, R.;
`
Frere, J. M.; Galleni, M.; Rossolini, G. M.; Dideberg, O.
J. Mol. Biol. 2003, 325, 651.
9. Sun, Q.; Law, A.; Crowder, M. W.; Geysen, H. M. Bioorg.
Med. Chem. Lett. 2006, 16, 5169.
10. Walter, M. W.; Felici, A.; Galleni, M.; Soto, R. P.;
16. Kemperman, G. J.; Roeters, T. A.; Hilberink, P. W. Eur.
J. Org. Chem. 2003, 9, 1681.
17. Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor,
R. J. Mol. Biol. 1997, 267, 727.
`
Adlington, R. M.; Baldwin, J. E.; Frere, J. M.; Gololobov,
M.; Schofield, C. J. Bioorg. Med. Chem. Lett. 1996, 6,
2455.
11. Walter, M. W.; Valladares, M. H.; Adlington, R. M.;
18. Olsen, L.; Pettersson, I.; Hemmingsen, L.; Adolph, H. W.;
Jorgensen, F. S. J. Comput. Aided Mol. Des. 2004, 18, 287.
19. Chevrier, B.; D’Orchymont, H.; Schalk, C.; Tarnus, C.;
Moras, D. Eur. J. Biochem. 1996, 237, 393.
`
Amicosante, G.; Baldwin, J. E.; Frere, J. M.; Galleni, M.;