The identification of new metallo-β-lactamase inhibitor leads from fragment-based screening

Article abstract of DOI:10.1016/j.bmcl.2011.04.027

The emergence of metallo-β-lactamases (MBLs) capable of hydrolysing a broad spectrum of β-lactam antibiotics is particularly concerning for the future treatment of bacterial infections. This work describes the discovery of lead compounds for the development of new inhibitors using a competitive colorimetric assay based on the chromogenic cephalosporin CENTA, and a 500 compound Maybridge library suitable for fragment-based screening. The interactions between identified inhibitory fragments and the active site of the MBL from Klebsiella pneumoniae and Pseudomonas aeruginosa were probed by in silico docking studies.

Full text of DOI:10.1016/j.bmcl.2011.04.027

Bioorganic & Medicinal Chemistry Letters 21 (2011) 3282–3285
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The identification of new metallo-b-lactamase inhibitor leads from
fragment-based screening
a
a
a
a
b
c
´
Peter Vella , Waleed M. Hussein , Eleanor W. W. Leung , Daniel Clayton , David L. Ollis , Nataša Mitic ,
Gerhard Schenk a,c, , Ross P. McGeary
a,d,
⇑
⇑
a The University of Queensland, School of Chemistry and Molecular Biosciences, Brisbane, QLD 4072, Australia
b The Australian National University Research School of Chemistry, Canberra, ACT 0200, Australia
c National University of Ireland-Maynooth, Department of Chemistry, Maynooth, Co. Kildare, Ireland
d The University of Queensland, School of Pharmacy, Brisbane, QLD 4072, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 22 February 2011
Revised 1 April 2011
Accepted 7 April 2011
Available online 13 April 2011
The emergence of metallo-b-lactamases (MBLs) capable of hydrolysing a broad spectrum of b-lactam
antibiotics is particularly concerning for the future treatment of bacterial infections. This work describes
the discovery of lead compounds for the development of new inhibitors using a competitive colorimetric
assay based on the chromogenic cephalosporin CENTA, and a 500 compound Maybridge™ library suitable
for fragment-based screening. The interactions between identified inhibitory fragments and the active
site of the MBL from Klebsiella pneumoniae and Pseudomonas aeruginosa were probed by in silico docking
studies.
Keywords:
Antibiotics resistance
Fragment-based screening
Molecular docking
Inhibition assays
Ó 2011 Elsevier Ltd. All rights reserved.
Metallo-b-lactamases
b-Lactam-based antibiotics are the gold standard for treating
bacterial infections. However, bacteria have now developed resis-
tance to these drugs through mechanisms such as changes in
membrane permeability, efflux pumps and the expression of lacta-
mase enzymes specialising in the destruction of b-lactam antibiot-
ics through hydrolysis.1,2 Of particular concern is the emergence of
the metallo-b-lactamases (MBLs), which are a family of zinc con-
taining metalloenzymes capable of hydrolysing a very broad range
of common b-lactam antibiotics, including penams (e.g., benzyl-
penicillin), cephalosporins (e.g., cephalothin, cefoxitin) and carba-
penems (e.g., meropenem and imipenem).3 Examples of MBLs
include VIM-2 and IMP-1, both of which have been isolated follow-
ing outbreaks of bacterial resistance in hospitals, and both of which
now have many documented isoforms.4–9 MBLs are not inhibited
by clavulanic acid, a drug commonly co-administered with b-lac-
tam antibiotics as an inhibitor of other types of lactamases.10
The main characteristics of MBLs are the presence of two Zn(II)
ions in the active site (although in some cases only one may be
(Fig. S1 in Supplementary data).11–14 The metal ions are proposed
to assist in the binding of the antibiotic substrate and the genera-
tion of an attacking nucleophile.14,15 MBLs are classified as B1, B2
or B3 type depending on the residues ligating the metal ions in
the active site (Figs. S1 and S2 in Supplementary data). A standard
numbering scheme for residues in MBLs has been developed based
on sequence alignments for B1, B2 and B3 MBLs, and this scheme is
used throughout this Letter.16 Various catalytic mechanisms have
been proposed for MBLs, and were summarised in a recent re-
view.15 Although the proposed mechanistic strategies vary with re-
spect to the role of the metal ions, the number of metal ions in the
active site, the mode of substrate binding and the identity of the
rate-limiting step, it is generally accepted that a nucleophilic at-
tack by a metal ion-bound hydroxide onto the antibiotic’s four-
membered lactam ring triggers ring opening and the hydrolysis
of the antibiotic to a therapeutically inactive form.
The imipenemase-1 (IMP-1) MBL from Pseudomonas aeruginosa
and its isoforms are documented in many outbreaks of antibiotic
resistant bacteria in medical facilities worldwide.17 Examples of
serious infections involving P. aeruginosa include bacteriaemia,
pneumonia, urosepsis and wound infections.18 Although it is capa-
ble of hydrolysing penams (e.g., penicillin), IMP-1 is particularly
efficient at hydrolysing cephalosporins and carbapenems.19 Mobile
genetic elements such as plasmids encoding MBLs allow resistance
to spread between unrelated bacterial species. It is therefore not
needed for catalytic activity) and an overall
abba protein fold
⇑
Corresponding authors. Tel.: +61 7 3365 4144; fax: +61 7 3365 4273 (G.S.); tel.:
+61 7 3365 3955; fax: +61 7 3346 3249 (R.P.M.).
E-mail addresses: schenk@uq.edu.au (G. Schenk), r.mcgeary@uq.edu.au (R.P.
McGeary).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.04.027

P. Vella et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3282–3285
3283
surprising that IMP-1 is also found in other bacterial species such
as Klebsiella pneumoniae.4,20 The ability for MBLs such as IMP-1
to bestow broad-spectrum antibiotic resistance to bacterial patho-
gens, transfer resistance genes between bacterial species, the lack
of a clinically useful inhibitor and the diminishing number of treat-
ment options available for use against resistant bacterial infections
highlights the need for the development of potent inhibitors
against such MBLs.21,22
moiety and are both known to inhibit MBL activity, however
L
-cap-
topril is commercially available and was thus used for the assays.
The inhibition of IMP-1 by -captopril was assessed using both a
L
plate reader and a cuvette assay using benzylpenicillin as a sub-
strate (Figs. S3a and S3b in Supplementary data). Interestingly,
CENTA was not a suitable substrate since captopril appeared to ini-
tiate its hydrolysis (data not shown). The mode of inhibition is
competitive with
a Kic of 12.5 2.4 lM (cuvette assay) and
Fragment-based screening has proved to be an effective strat-
egy for the discovery and development of drug leads.23–25 In this
approach, low molecular weight ligands that bind to proteins of
interest with weak binding affinities (typically around 1 mM) are
identified. NMR spectroscopy or X-ray crystallography are subse-
quently used to determine the location and three-dimensional
binding modes of these fragments. Both competitive and non-com-
petitive binders can be used as starting points for further drug lead
development, by tethering fragments together to develop more po-
tent ligands, and by designing drug-like properties into these com-
pounds.26 Successes using this approach include the development
of a CDK2 inhibitor now in clinical trials by Astex Therapeutics
and the development of a Bcl-2 inhibitor by Abbott Laborato-
ries.27,28 A fragment based screening approach based around the
chromogenic cephalosporin CENTA (see 15, Scheme 1 in the Sup-
plementary data) and the 500-compound Maybridge™ R03 frag-
ment library was therefore devised here to identify new
inhibitors against IMP-1.29,30 Although this approach was expected
to initially yield compounds with inhibition constants only in the
7.2 1.2 M (96-well plate assay) demonstrating that the 96-well
l
high-throughput assay provides reliable kinetic data. In the initial
fragment screening assays the percent inhibition was determined
by comparing the rate of CENTA turnover between assays where
the fragment was present (at 1 mM concentration) or absent. The
inhibition cut-off was set at 50%, with fragments meeting this cri-
terion being further analysed using inhibition assays as described
for L-captopril (vide supra and Figs. S3a and S3b in the Supplemen-
tary data). The remaining fragments were not further assessed.
Ten compounds with inhibition constants within the high
micromolar to low millimolar range were thus identified and their
inhibition constants, Ki, and mode of inhibition were assessed
(compounds 1–10, Fig. 1). The inhibition constants are summarised
in Table 1, and a plot of the Ki determination for the most potent
compound, fragment 1, appears in the Supplementary data
(Fig. S4). Fragment 1 also has the highest ligand efficiency of the
compounds tested.34 Although the compounds generally exhibited
mixed inhibition (both competitive and uncompetitive inhibition
modes were observed), the uncompetitive mode of inhibition
was usually dominant. Only fragment 10 was observed to inhibit
MBL activity in a predominantly competitive way. As the aim of
a fragment based screen is to identify small molecules that may
form a part of a larger, more potent inhibitor, the discovery of a
number of uncompetitively bound fragments is encouraging for
the development of lead compounds to inhibit MBLs. The uncom-
petitive mode of inhibition indicates that the fragment does not
prevent the substrate CENTA from binding to the active site; in-
stead, a ternary enzyme–substrate–inhibitor (ESI) complex is
formed that does not promote hydrolysis of the substrate. This en-
zyme–substrate–inhibitor (ESI) may provide the basis for the syn-
thesis of potent and specific IMP-1 inhibitors. To gain insight into
its possible structure, in silico docking was employed. The aim
was to dock the substrate, CENTA, and a fragment into the crystal
structure of IMP-1 (PDB code: 1JJT).35 The fragments selected for
docking were 1 (strongest biding affinity), 6 (an uncompetitive
inhibitor fragment) and 10 (a competitive inhibitor fragment) (Ta-
low mM to high lM range, the philosophy of the fragment-based
approach is to first identify, and then develop such compounds into
more potent inhibitors by synthetic modifications guided by in sil-
ico docking or crystallography studies.31,32 The synthesis of CENTA,
assays used, the in silico docking approach and other experimental
procedures employed in this study are detailed in the Supplemen-
tary data.
The aim of the present study was to identify new inhibitors of
MBLs using a fragment-based screening approach. As a target we
selected IMP-1 from P. aeruginosa and K. pneumoniae as these
pathogens have already displayed significant resistance to many,
if not most, commonly used antibiotics. To search for new inhibi-
tors of IMP-1, the 500 compound Maybridge™ R03 fragment li-
brary was screened. A 96-well plate assay was used to facilitate
rapid throughput in the analysis. Captopril, a known inhibitor of
MBLs was used to test the sensitivity of this assay.33 The D- and
L-diastereomers of captopril differ in the chirality of the proline
Figure 1. L-Captopril (11) and fragments 1–10 that were identified from the Maybridge™ library as potential leads to design and synthesise novel MBL inhibitors.

3284
P. Vella et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3282–3285
Table 1
Inhibition constants for L-captopril and the ten most potent fragment compounds
from the Maybridge™ library
Number
Identifier
Kiuc (mM)
Kic (mM)
1
2
3
4
5
6
7
8
9
mo07352
btb7340
cc18509
tl01011
cc43214
cc43309
cc39814
cc43314
km01548
sb00671
Captopril
0.41 0.10
0.43 0.07
0.50 0.09
0.68 0.17
0.78 0.14
0.79 0.10
0.90 0.17
1.18 0.23
1.20 0.30
—
0.89 1.36
0.99 0.87
—
0.76 0.45
1.64 1.60
—
0.51 0.27
1.51 1.08
—
10
11
0.97 0.60
—
0.0125 0.0024a
–Indicates result >2 mM.
a
Substrate used in the assays was penicillin G, cuvette assay.
ble 1). As the residues Glu 60, Val 61 and Trp 64 form part of a flex-
ible loop on IMP-1, the docking was run with these residues
marked as flexible. The reliability of the docking method was
tested using the competitive inhibitor
-Captopril docked into the active site of IMP-1 (Fig. S5) shows a
similar binding mode to that observed for crystal structures of
D-captopril as a probe.
D
MBLs that have D-captopril bound, such as the B1-type MBL BlaB
(PDB code: 1M2X)36 where the sulfur atom of the inhibitor coordi-
nates in a bridging mode to both metal ions in the active site. Since
D-captopril occupies the space close to the metal ions it effectively
competes with the substrate for a spot in the immediate vicinity of
the catalytic centre, a binding mode that is consistent with com-
petitive inhibition. Subsequently, the substrate CENTA and the
most potent of the fragment compounds, 1, were docked into the
active site of IMP-1, first separately, then as an ESI complex. Frag-
ment 1 docks into the active site of IMP-1 with two plausible con-
formations. In the first conformation, the N of the amino group of 1
occupies a position close to both metal ions in the active site (Zn1:
4.0 Å; Zn2: 3.2 Å; Fig. 2a). In the second conformation the fragment
Figure 2. (a) Surface view of the IMP-1 active site for the highest Autodock Vina
score conformation of IMP-1 with fragment 1 docked into the active site in a
proposed ‘competitive mode’. (b) Fragment 1 docked into the active site in a
proposed ‘un-competitive mode’. (c) CENTA substrate docked into the active site.
For clarity, Trp 64 and Glu 60 of the flexible loop adjacent to the active site has been
given a ‘stick’ representation. Atom colours are as follows: blue—nitrogen, red—
oxygen, white—carbon (on IMP-1), green—carbon (on inhibitor or CENTA), gold—
sulfur, purple—zinc active site metals.
interacts through p-stacking with residue Trp 64, which is located
on the flexible loop close to the active site of IMP-1 (Fig. 2b). As the
inhibition data (Table 1) indicate fragment 1 is a mixed-type inhib-
itor, we thus assign the first, lower energy conformation as the
competitive mode of binding and the second as the uncompetitive
one. Consistent with the uncompetitive binding mode of fragment
6, the docking solution suggests that this fragment does not bind to
the zinc ions within the active site (Fig. S6). However, the docking
shows fragment 6 interacting through hydrogen bonding with Lys
224 (through the N atom on the imidazole ring) and Ser 119 of the
enzyme (through the hydroxyl group). The docking solution for
fragment 10 suggests that it closely binds a zinc ion in the active
site using two nitrogen atoms from the triazole ring (Zn1: 2.6,
2.5 Å; Zn2: 2.5 Å), consistent with the competitive mode of inhibi-
tion observed from the kinetic assays (Fig. S7).
The highest ranked docking solution of the substrate CENTA
docked into IMP-1 is shown in Figure 2c. CENTA docks into the ac-
tive site with the nitrogen of the lactam ring positioned directly in
front of the active site metal ions, thus being in a position that
facilitates a nucleophilic attack by a metal ion-bound hydroxide.
The carbonyl group on the lactam ring of CENTA docks closely to
Zn1 and the carboxylic acid group is shown interacting with Lys
224. The observation of this particular interaction through in silico
docking is encouraging as previous work by Haruta et al. have
demonstrated that substitution of this lysine residue to an argi-
nine, alanine or glutamic acid in IMP-1 results in an increased
Km, and thus weaker substrate binding.37 Furthermore, they
hypothesised that the positive charge of the lysine was important
for the binding of the carboxylic acid group of b-lactam antibiotic
substrates, and that this interaction was one possible explanation
for why MBLs generally lack activity towards b-lactam monobac-
tam antibiotics, which lack this particular carboxylic acid group.
Since the ESI complex may point towards the design of more
potent next generation inhibitors for MBLs, docking was used to
model the ESI complex by docking CENTA into IMP-1 with frag-
ment 1 already bound (Fig. S8 in Supplementary data), and a super-
position of CENTA (Fig. 2c) and compound 1 in the uncompetitive
binding mode (Fig. 2b) docked into IMP-1 was carried out. While
the docking approach does not account for all protein dynamics
during substrate and/or inhibitor binding, the superimposed struc-
ture does show that fragment 1 and CENTA are in close proximity
near the active site, and are thus amenable to linking together to
form a putative, more potent competitive inhibitor of IMP-1
(Fig. S9 in Supplementary data). For example, extending function-
ality from the methyl ester may allow occupation of a pocket that
is present to the left of Lys 224, and additional pendant groups
could be attached to the phenyl ring to bind the Zn(II) metals with-
in the active site.
In conclusion, a number of compounds with sub-millimolar
inhibition constants (both competitive and non-competitive
modes of binding) from the Maybridge™ R03 fragment library,
have been identified as inhibitors of the IMP-1 MBL. This work also
confirms the use of the chromogenic cephalosporin CENTA as a
suitable substrate and alternative to nitrocefin for the purpose of

P. Vella et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3282–3285
3285
14. Crowder, M. W.; Spencer, J.; Vila, A. J. Acc. Chem. Res. 2006, 39, 721.
inhibitor screening. The identification of ten inhibitor leads which
predominantly show an uncompetitive mode of inhibition pro-
vides a platform for the further development of potent compounds
against IMP-1 in particular, and MBLs in general.
15. Page, M. I.; Badarau, A. Bioinorg. Chem. Appl. 2008, 2008, Article ID 576297.
16. Galleni, M.; Lamotte-Brasseur, J.; Rossolini, G. M.; Spencer, J.; Dideberg, O.;
Frère, J.-M. Antimicrob. Agents Chemother. 2001, 45, 660.
17. Peleg, A. Y.; Franklin, C.; Bell, J.; Spelman, D. W. J. Antimicrob. Chemother. 2004,
54, 699.
18. Kerr, K. G.; Snelling, A. M. J. Hosp. Infect. 2009, 73, 338.
19. Laraki, N.; Franceschini, N.; Rossolini, G. M.; Santucci, P.; Meunier, C.; de Pauw,
E.; Amicosante, G.; Frère, J. M.; Galleni, M. Antimicrob. Agents Chemother. 1999,
43, 902.
20. Penteado, A. P.; Castanheira, M.; Pignatari, A. C. C.; Guimarães, T.; Mamizuka, E.
M.; Gales, A. C. Diagn. Microbiol. Infect. Dis. 2009, 63, 87.
Acknowledgement
This work was supported by the National Health and Medical
Research Council of Australia Project Grant (631443).
21. Li, J.; Nation, R. L.; Turnidge, J. D.; Milne, R. W.; Coulthard, K.; Rayner, C. R.;
Paterson, D. L. Lancet Infect. Dis. 2006, 6, 589.
22. Livermore, D. M. Clin. Infect. Dis. 2002, 34, 634.
Supplementary data
23. Fragment-based Approaches in Drug Discovery; Jahnke, W., Erlanson, D. A., Eds.;
Wiley-VCH: Weinheim, 2006.
24. Fragment-based Drug Discovery: A Practical Approach; Zartler, E. R., Shapiro, M.
J., Eds.; John Wiley & Sons: Chichester, UK, 2008.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.04.027.
25. Zartler, E. R.; Shapiro, M. J. Curr. Opin. Chem. Biol. 2005, 9, 366.
26. Chessari, G.; Woodhead, A. J. Drug Discovery Today 2009, 14, 668.
27. Wyatt, P. G.; Woodhead, A. J.; Berdini, V.; Boulstridge, J. A.; Carr, M. G.; Cross, D.
M.; Davis, D. J.; Devine, L. A.; Early, T. R.; Feltell, R. E.; Lewis, E. J.; McMenamin,
R. L.; Navarro, E. F.; O’Brien, M. A.; O’Reilly, M.; Reule, M.; Saxty, G.; Seavers, L.
C. A.; Smith, D.-M.; Squires, M. S.; Trewartha, G.; Walker, M. T.; Woolford, A. J.
A. J. Med. Chem. 2008, 51, 4986.
28. Shuker, S. B.; Hajduk, P. J.; Meadows, R. P.; Fesik, S. W. Science 1996, 274, 1531.
29. Bebrone, C.; Moali, C.; Mahy, F.; Rival, S.; Docquier, J. D.; Rossolini, G. M.;
Fastrez, J.; Pratt, R. F.; Frère, J.-M.; Galleni, M. Antimicrob. Agents Chemother.
2001, 45, 1868.
30. Thermo Fisher Scientific Inc. 2010, http://www.maybridge.com/Images/pdfs/
MB_Ro3_fragment_flyer_2010_EEM_Final.pdf (accessed January 2011).
31. Hartshorn, M. J.; Murray, C. W.; Cleasby, A.; Frederickson, M.; Tickle, I. J.; Jhoti,
H. J. Med. Chem. 2004, 48, 403.
32. Chen, Y.; Shoichet, B. K. Nat. Chem. Biol. 2009, 5, 358.
33. Heinz, U.; Bauer, R.; Wommer, S.; Meyer-Klaucke, W.; Papamichaels, C.;
Bateson, J.; Adolph, H.-W. J. Biol. Chem. 2003, 278, 20659.
References and notes
1. Masuda, N.; Sakagawa, E.; Ohya, S. Antimicrob. Agents Chemother. 1995, 39, 645.
2. Bornet, C.; Chollet, R.; Malléa, M.; Chevalier, J.; Davin-Regli, A.; Pagés, J.-M.;
Bollet, C. Biochem. Biophys. Res. Commun. 2003, 301, 985.
3. Felici, A.; Amicosante, G.; Oratore, A.; Strom, R.; Ledent, P.; Joris, B.; Fanuel, L.;
Frère, J. M. Biochem. J. 1993, 291, 151.
4. Fukigai, S.; Alba, J.; Kimura, S.; Iida, T.; Nishikura, N.; Ishii, Y.; Yamaguchi, K. Int.
J. Antimicrob. Agents 2007, 29, 306.
5. Herbert, S.; Halvorsen, D. S.; Leong, T.; Franklin, C.; Harrington, G.; Spelman, D.
Infect. Control Hosp. Epidemiol. 2007, 28, 98.
6. Gibb, A. P.; Tribuddharat, C.; Moore, R. A.; Louie, T. J.; Krulicki, W.; Livermore, D.
M.; Palepou, M.-F. I.; Woodford, N. Antimicrob. Agents Chemother. 2002, 46, 255.
7. Ryoo, N. H.; Lee, K.; Lim, J.-B.; Lee, Y. H.; Bae, I. K.; Jeong, S. H. Diagn. Microbiol.
Infect. Dis. 2009, 63, 115.
8. Crespo, M. P.; Woodford, N.; Sinclair, A.; Kaufmann, M. E.; Turton, J.; Glover, J.;
Velez, J. D.; Castaneda, C. R.; Recalde, M.; Livermore, D. M. J. Clin. Microbiol.
2004, 42, 5094.
9. Lahey Clinic. http://www.lahey.org/studies (accessed January 2011).
10. Ohsuka, S.; Arakawa, Y.; Horii, T.; Ito, H.; Ohta, M. Antimicrob. Agents Chemother.
1995, 39, 1856.
11. Paul-Soto, R.; Bauer, R.; Frère, J.-M.; Galleni, M.; Meyer-Klaucke, W.; Nolting,
H.; Rossolini, G. M.; de Seny, D.; Hernandez-Valladares, M.; Zeppezauer, M.;
Adolph, H.-W. J. Biol. Chem. 1999, 274, 13242.
34. Schultes, S.; de Graaf, C.; Haaksma, E. E. J.; de Esch, I. J. P.; Leurs, R.; Krämer, O.
Drug Discovery Today: Technol. 2010, 7, E157.
35. Toney, J. H.; Hammond, G. G.; Fitzgerald, P. M. D.; 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.
´
36. Garcıa-Sáez, I.; Isabel; Hopkins, J.; Papamicael, C.; Franceschini, N.;
Amicosante, G.; Rossolini, G. M.; Galleni, M.; Frère, J. M.; Jean-Marie;
Dideberg, O. J. Biol. Chem. 2003, 278, 23868.
37. Haruta, S.; Yamamoto, E. T.; Eriguchi, Y.; Sawai, T. FEMS Microbiol. Lett. 2001,
197, 85.
12. Bebrone, C. Biochem. Pharmacol. 2007, 74, 1686.
13. Morán-Barrio, J.; González, J. M.; Lisa, M. N.; Costello, A. L.; Peraro, M. D.;
Carloni, P.; Bennett, B.; Tierney, D. L.; Limansky, A. S.; Viale, A. M.; Vila, A. J. J.
Biol. Chem. 2007, 282, 18286.