Selective trihydroxyazepane NagZ inhibitors increase sensitivity of Pseudomonas aeruginosa to β-lactams

Article abstract of DOI:10.1039/c3cc46646a

AmpC β-lactamase confers resistance to β-lactam antibiotics in many Gram negative bacteria. Inducible expression of AmpC requires an N-acetylglucosaminidase termed NagZ. Here we describe the synthesis and characterization of hydroxyazepane inhibitors of N

Full text of DOI:10.1039/c3cc46646a

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Selective trihydroxyazepane NagZ inhibitors increase
sensitivity of Pseudomonas aeruginosa to b-lactams†
Cite this: Chem. Commun., 2013,
49, 10983
Martine Mondon,a Soo Hur,b Grishma Vadlamani,c Prerana Rodrigues,c
Received 30th August 2013,
Polina Tsybina,b Antonio Oliver,d Brian L. Mark,c David J. Vocadlo*b and
Accepted 25th September 2013
a
´
Yves Bleriot*
DOI: 10.1039/c3cc46646a
www.rsc.org/chemcomm
AmpC b-lactamase confers resistance to b-lactam antibiotics in many
Gram negative bacteria. Inducible expression of AmpC requires an
N-acetylglucosaminidase termed NagZ. Here we describe the synthesis
and characterization of hydroxyazepane inhibitors of NagZ.
We find that these inhibitors enhance the susceptibility of clinically
relevant Pseudomonas aeruginosa to b-lactams.
Fig. 1 NagZ substrate, product (GlcNAc), PUGNAc, and azepane 1. The stem
peptide is not shown for reasons of clarity.
Resistance to frontline b-lactams in Gram negative bacteria is a
growing clinical problem.1 Given the prevalence of bacterial
pathogens that harbour either inducible chromosomal AmpC
NagZ enzymes belong to the GH3 family of the glycoside
or inducible plasmid-borne AmpC,2–5 new strategies to sup- hydrolases.11 Detailed kinetic12,13 and X-ray structural studies8,14,15
press this resistance mechanism are needed.
of NagZ enzymes have revealed that they use a two step mechanism
Inducible AmpC expression is regulated by peptidoglycan meta- involving the formation and breakdown of a covalent glycosyl-
bolites produced within the bacterium.6 The presence of b-lactams enzyme intermediate. The 2-acetamido group of the GlcNAc residue
leads to increased cytosolic levels of the metabolite GlcNAc-1,6- of the substrate plays no role in catalysis, which is in contrast to
anhydroMurNAcpeptide, which is cleaved by a conserved enzyme of human b-acetylglucosaminidases from GH2013,16 and GH84,17,18
the Gram-negative peptidoglycan cell wall recycling pathway termed which use a catalytic mechanism in which the 2-acetamido group
NagZ (Fig. 1).7 NagZ removes the non-reducing GlcNAc residue to of the substrate acts as a nucleophile to yield a bicyclic oxazoline
liberate 1,6-anhydroMurNAc peptides, which induce transcription intermediate. b-Hexosaminidase A (HexA) and b-hexosaminidase B
of AmpC b-lactamase to confer resistance to the b-lactam anti- (HexB) are GH20 enzymes and their dysfunction gives rise to
biotic.7 b-Lactam therapy also frequently selects for mutations heritable lysosomal storage diseases. The GH84 enzyme known as
that cause hyperproduction of AmpC. One approach that is O-GlcNAcase (OGA) removes O-linked b-N-acetylglucosamine residues
being explored to reverse resistance to b-lactams is to generate (O-GlcNAc) from serine and threonine residues of nucleocytoplasmic
inhibitors of NagZ that can act to hinder the production of proteins.19,20 Examination of the active sites of the GH3, GH20, and
AmpC by disrupting its induction pathway.8–10 One challenge GH84 enzymes reveals that the active site topology interacting with
facing this approach is that NagZ inhibitors can also inhibit the 2-acetamido group of the substrate differs.21 For GH20 and GH84
functionally related human enzymes.
enzymes, this pocket envelops the acetamido group, whereas for the
NagZ enzymes this substituent rests against the surface of the enzyme
and is solvent exposed. Exploiting this discrepancy by incorporating
N-acyl substituents of various volumes therefore offers a tractable
route to generating selective inhibitors of NagZ.
a
´
´
´
Universite de Poitiers, IC2MP, UMR CNRS 7285, Equipe ‘‘Synthese Organique’’
Groupe Glycochimie, 4 rue Michel Brunet, 86022 Poitiers Cedex, France.
E-mail: yves.bleriot@univ-poitiers.fr
b Department of Chemistry, Simon Fraser University, 8888 University Drive,
Burnaby, British Columbia, Canada V5S 1P6. E-mail: dvocadlo@sfu.ca
c Department of Microbiology, University of Manitoba, Buller Building, Winnipeg,
Several different glycosidase inhibitors that can inhibit diverse
exo-N-acetyl-b-glucosaminidases, including various enzymes from
GH3, GH20, and GH84, have been reported. Among the most
potent of these inhibitors are PUGNAc,22 the iminosugar-based
AcDNJ,8,10 and the glycosylimidazoles.23 An additional class that
has gained attention over the last two decades is the family of
polyhydroxylated azepanes.24–28
Manitoba, Canada R3T 2N2
d
´
Servicio de Microbiologıa, Hospital Son Espases, Ctra Valldemossa 79,
07010 Palma de Mallorca, Spain
† Electronic supplementary information (ESI) available: Experimental methods
and structure parameters. See DOI: 10.1039/c3cc46646a
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 10983--10985 10983

View Article Online
Communication
ChemComm
Table 1 Inhibition of Salmonella typhimurium NagZ, OGA, and HexA by azepanes
1 and 9–16
Compound
1
9
10
11
12
15
16
OGA Ki (mM)
NagZ Ki (mM)
HexA Ki (mM)
(Ki OGA/Ki NagZ)
(Ki HexA/Ki NagZ)
0.7
0.4
3.6
1.8
2.1
2.2
0.4
NI
5.5
+
47
7.4
NI
6.7
+
190
27
NI
7.1
+
NI
NI
NI
NA
NA
22
NI
51
NA
NA
2.0
NI
14
NA
NA
NI: no significant inhibition at 500 mM, NA: not applicable, italic
numbers Ki by Dixon analysis, +: estimated to be at least >1000. For
errors on these measurements see Table S2 (ESI).
Scheme 1 Synthesis of 3-N-acyl azepanes 1 and 9–12.
When protonated, the basic nitrogen of azepanes is thought to
confer resemblance to the oxocarbenium ion-like transition state
proposed for NagZ and other glycoside hydrolases. The azepanes
also display increased conformational flexibility compared to the six-
membered iminosugar inhibitors.29 This increased plasticity enables
these molecules to be accommodated within the active sites of
glycoside hydrolases, as illustrated by the (3S,4R,5R,6S)-3-acetamido-
4,5,6-trihydroxyazepane 1, that mimics GlcNAc and inhibits eukar-
yotic GH20 and GH84 enzymes with Ki values ranging from 0.4 mM
to 11 mM.26 The versatility of the azepane scaffold prompted us to
explore the potential of various (3S,4R,5R,6S)-3-N-acyl-4,5,6-trihydrox-
yazepanes as inhibitors of NagZ.
We find that azepane 1 is a good inhibitor of NagZ but shows
poor selectivity (Table 1). To gain insight into the molecular basis of
NagZ inhibition by azepane 1 we determined the X-ray crystal
structure of this molecule bound to NagZ from the opportunistic
human pathogen Burkholderia cenocepacia (BcNagZ). The complex
was solved by molecular replacement using a model of BcNagZ
(PDB 4GNV) from which ligands and solvents were removed. Like
other NagZ enzymes BcNagZ is a single-domain enzyme containing
a (b/a)8 barrel fold (TIM barrel) (Fig. S1, ESI†). The structure of this
enzyme in complex with 1 reveals several hydrogen-bonding inter-
actions with the enzyme that contribute to the potency of this
molecule as an inhibitor of GH3 NagZ enzymes (Fig. 2A).
surrounding the endocyclic nitrogen as well as the acetyl group.
Based on these structures, and the structures of GH84 OGA homo-
logues as well as GH20 human hexosaminidase, we envisioned that
selectivity toward NagZ could be obtained by exploring structural
variations at the acetamido group and through alkylation of the
endocyclic nitrogen. We therefore generated a panel of 3-N-acyl
derivatives of (3S,4R,5R,6S)-3-acetamido-4,5,6-trihydroxyazepane 1
and evaluated these against NagZ, OGA, and HexA.
We modified the route to 130a by using PPh3 on a solid support.
It enabled smooth conversion of the starting azidolactol into the
corresponding N,O acetal. Its reduction with NaBH3CN followed
by protection of the endocyclic nitrogen as its benzyl carbamate
afforded 3-hydroxyazepane 2 in 45% yield. Activation of the free OH
as its mesylate followed by azide displacement furnished azidoaze-
pane 3 in 75% yield. Reduction with PPh3 followed by acylation of
the resulting crude amine with acyl anhydrides or acyl chlorides
furnished the corresponding 3-N-acyl azepanes 4–8 in 45–74% yield.
Hydrogenolysis yielded target azepanes 1 and 9–12 (Scheme 1).
The azepane inhibitor scaffold was also elaborated by endo-
cyclic N-alkylation. Such modified polyhydroxylated piperidines can
be selective and potent inhibitors of various glycosidases, including
family GH20 b-glucosaminidases.30b We therefore examined
such modifications on the selectivity and potency of azepanes for
these enzymes. Such N-alkylation of azasugars mimics the aglycon
moiety of substrates of these enzymes.31 Here we investigated the
commonly seen n-butyl group and a phenylbutyl group that we
intended to mimic the aglycon phenylcarbamate substituent of
PUGNAc. Azepane 4 was hydrogenolyzed under mild conditions to
selectively remove the Cbz group and N-alkylated using butyl
bromide or phenylbutyl bromide in the presence of K2CO3 to yield
azepanes 13 (80% yield) and 14 (64% yield) respectively. Hydro-
genolysis under acidic conditions furnished the corresponding
trihydroxyazepanes 15 and 16 (Scheme 2).
In particular, the catalytic nucleophile of BcNagZ, Asp253, forms
a 2.7 Å hydrogen-bond with the endocyclic nitrogen (N1) of 1—an
interaction that cannot occur between the nucleophile and GlcNAc,
(Fig. 2B) suggesting this interaction of 1 underlies its potency
against NagZ. The structures also reveal significant space
With this set of compounds in hand we evaluated their
inhibition of NagZ, OGA, and HexA. We find that increasing the
Fig. 2 Crystal structure of BcNagZ bound to azepane 1. (A) Active site residues
(grey) and 1 (yellow) are drawn as sticks. The NagZ consensus sequence (blue) has
the putative histidine general acid–base in sticks. Hydrogen bonding interactions
are represented by dashed magenta lines. Electron density is an Fo–Fc map
(3.0 rmsd). (B) Superimposition of the BcNagZ–1 complex with that of BcNagZ
bound to GlcNAc (cyan). See Table S1 (ESI†) for crystal and refinement statistics.
Scheme 2 Synthesis of N-alkyl azepanes 15 and 16.
c
10984 Chem. Commun., 2013, 49, 10983--10985
This journal is The Royal Society of Chemistry 2013

View Article Online
ChemComm
Communication
size of the N-acyl group in the series including 1, 9–11 has a structure factors for the BcNagZ–azepane complex are available in
modest effect with regard to inhibition of NagZ and OGA in that the Protein Data Bank (4MSS).
the Ki values increase with the bulk of this N-acyl group by
Notes and references
approximately 30-fold for NagZ and 200-fold for OGA. Inter-
estingly, branching of the aliphatic group, as in compound 12,
1 J. M. Thomson and R. A. Bonomo, Curr. Opin. Microbiol., 2005, 8, 518–524.
leads to a loss of inhibition against all these enzymes. As
previously observed for other inhibitors, increasing the bulk
of the N-acyl group beyond acetyl leads to a dramatic loss of
potency against HexA as seen for compound 9, which differs in
structure from 1 by only one methylene unit.
Because N-alkylation of iminosugars has often resulted in
significantly enhanced binding to glycoside hydrolases, including
to OGA, we speculated that N-alkylated azepanes 15 and 16 would
be more potent than 1 against these enzymes. Surprisingly,
however, we find that both these compounds abolished inhibition
against NagZ and impaired binding to OGA by a modest 2–30 fold
and to HexA by 4–15 fold.
Since some of these azepanes are good inhibitors of NagZ, in
particular 10, which retained fair potency and now showed some
(7-fold) selectivity for NagZ over both OGA and Hex, we decided to
explore their potential to block NagZ in bacteria and attenuate
inducible AmpC-mediated b-lactam resistance within a relevant
bacterial model representative of clinically occurring P. aeruginosa.
Spontaneous inactivation of dacB, which encodes the non-essential
penicillin-binding protein 4 (PBP4), is the most common cause
of AmpC hyper-production and high-level b-lactam antibiotic
2 P. A. Bradford, C. Urban, N. Mariano, S. J. Projan, J. J. Rahal and
K. Bush, Antimicrob. Agents Chemother., 1997, 41, 563–569.
3 V. Miriagou, L. S. Tzouvelekis, L. Villa, E. Lebessi, A. C. Vatopoulos,
A. Carattoli and E. Tzelepi, Antimicrob. Agents Chemother., 2004, 48,
3172–3174.
4 G. Barnaud, G. Arlet, C. Verdet, O. Gaillot, P. H. Lagrange and
A. Philippon, Antimicrob. Agents Chemother., 1998, 42, 2352–2358.
5 R. Nakano, R. Okamoto, Y. Nakano, K. Kaneko, N. Okitsu, Y. Hosaka
and M. Inoue, Antimicrob. Agents Chemother., 2004, 48, 1151–1158.
6 S. Sobhanifar, D. T. King and N. C. Strynadka, Curr. Opin. Struct.
Biol., 2013, DOI: 10.1016/j.sbi.2013.07.008.
7 Q. Cheng, H. Li, K. Merdek and J. T. Park, J. Bacteriol., 2000, 182,
4836–4840.
8 K. A. Stubbs, M. Balcewich, B. L. Mark and D. J. Vocadlo, J. Biol.
Chem., 2007, 282, 21382–21391.
9 T. Yamaguchi, B. Blazquez, D. Hesek, M. Lee, L. I. Llarrull,
B. Boggess, A. G. Oliver, J. F. Fisher and S. Mobashery, ACS Med.
Chem. Lett., 2012, 3, 238–242.
10 K. A. Stubbs, J.-P. Bacik, G. E. Perley-Robsertson, G. E. Whitworth,
T. M. Gloster, D. J. Vocadlo and B. L. Mark, ChemBioChem, 2013, 14,
1973–1981.
11 B. L. Cantarel, P. M. Coutinho, C. Rancurel, T. Bernard, V. Lombard
and B. Henrissat, Nucleic Acids Res., 2009, 37, D233–D238.
12 D. J. Vocadlo, C. Mayer, S. He and S. G. Withers, Biochemistry, 2000,
39, 117–126.
13 D. J. Vocadlo and S. G. Withers, Biochemistry, 2005, 44, 12809–12818.
14 S. Litzinger, S. Fischer, P. Polzer, K. Diederichs, W. Welte and
C. Mayer, J. Biol. Chem., 2010, 285, 35675–35684.
resistance in P. aeruginosa.32 We therefore tested whether 1 or 15 J. P. Bacik, G. E. Whitworth, K. A. Stubbs, D. J. Vocadlo and
10 could potentiate the efficacy of the b-lactam ceftazidime
B. L. Mark, Chem. Biol., 2012, 19, 1471–1482.
16 B. L. Mark, D. J. Vocadlo, S. Knapp, B. L. Triggs-Raine, S. G. Withers
against a dacB null mutant of P. aeruginosa (Table S3, ESI†).
and M. N. James, J. Biol. Chem., 2001, 276, 10330–10337.
Neither 1 nor 10 had any effect on bacterial growth on their 17 Y. He, M. S. Macauley, K. A. Stubbs, D. J. Vocadlo and G. J. Davies,
J. Am. Chem. Soc., 2010, 132, 1807–1809.
18 M. S. Macauley, G. E. Whitworth, A. W. Debowski, D. Chin and
own. However, use of 1 mM 1 halved the minimum inhibitory
concentration (MIC) of P. aeruginosa DdacB from 24 mg mlÀ1
D. J. Vocadlo, J. Biol. Chem., 2005, 280, 25313–25322.
ceftazidime in the absence of the inhibitor to 12 mg mlÀ1
.
,
19 D. L. Dong and G. W. Hart, J. Biol. Chem., 1994, 269, 19321–19330.
20 Y. Gao, L. Wells, F. I. Comer, G. J. Parker and G. W. Hart, J. Biol.
Chem., 2001, 276, 9838–9845.
21 M. D. Balcewich, K. A. Stubbs, Y. He, T. W. James, G. J. Davies,
D. J. Vocadlo and B. L. Mark, Protein Sci., 2009, 18, 1541–1551.
22 M. Horsch, L. Hoesch, A. Vasella and D. M. Rast, Eur. J. Biochem.,
1991, 197, 815–818.
23 M. Terinek and A. Vasella, Helv. Chim. Acta, 2005, 88, 10–22.
24 S. D. Orwig, Y. L. Tan, N. P. Grimster, Z. Yu, E. T. Powers, J. W. Kelly
and R. L. Lieberman, Biochemistry, 2011, 50, 10647–10657.
25 H. Li, F. Marcelo, C. Bello, P. Vogel, T. D. Butters, A. P. Rauter, Y. Zhang,
Azepane 10 led to a greater 3-fold reduction in MIC to 8 mg mlÀ1
despite this derivative being 18-fold less potent in vitro relative to
azepane 1, suggesting that its uptake or diffusion into the cells
must occur more easily. Notably, this value meets the susceptibility
breakpoint of the Clinical and Laboratory Standards Institute
(CLSI) for ceftazidime.
The synthetic accessibility of polyhydroxy azepanes, their
stability and their ability to suppress b-lactam resistance in the
´
M. Sollogoub and Y. Bleriot, Bioorg. Med. Chem., 2009, 17, 5598–5604.
´
clinically relevant P. aeruginosa dacB null mutant suggest that these 26 F. Marcelo, Y. He, S. A. Yuzwa, L. Nieto, J. Jimenez-Barbero,
´
M. Sollogoub, D. J. Vocadlo, G. D. Davies and Y. Bleriot, J. Am. Chem.
compounds could prove useful. The availability of a NagZ–azepane
complex will facilitate the design of improved NagZ azepane inhi-
bitors. Finally, the recent observation that inactivation of NagZ
also blocks the emergence of mutations giving rise to b-lactam
resistance33 makes NagZ an enzyme of high interest; its inhibition
Soc., 2009, 131, 5390–5392.
27 (a) K. Dax, B. Gaigg, V. Grassberger, B. Kolblinger and A. E. Stu¨tz,
´
J. Carbohydr. Chem., 1990, 9, 479–499; (b) S. Pino-Gonzalez,
C. Assiego and N. Onas, Targets Heterocycl. Syst., 2004, 8, 300–330.
28 X. H. Qian, F. MorisVaras and C. H. Wong, Bioorg. Med. Chem. Lett.,
1996, 6, 1117–1122.
˜
˜
should not only block antibiotic resistance but could also hinder 29 K. Martinez-Mayorga, J. L. Medina-Franco, S. Mari, F. J. Canada,
´
E. Rodriguez-Garcia, P. Vogel, H. Q. Li, Y. Bleriot, P. Sina¨y and
development of mutations leading to b-lactam resistance.
´
J. Jimenez-Barbero, Eur. J. Org. Chem., 2004, 4119–4129.
We thank Dr Silvia Cardona for B. cenocepacia genomic DNA.
YB thanks ‘‘Vaincre la Mucoviscidose’’ for support. DJV and BLM
thank the Canadian Institutes of Health Research (MOP 97818) and
Cystic Fibrosis Canada for funding. BLM holds a Manitoba Research
´
´
´
30 (a) M. Mondon, N. Fontelle, J. Desire, F. Lecornue, J. Guillard,
´
J. Marrot and Y. Bleriot, Org. Lett., 2012, 14, 870–873; (b) C. W. Ho,
S. D. Popat, T. W. Liu, K. C. Tsai, M. J. Ho, W. H. Chen, A. S. Yang and
C. H. Lin, ACS Chem. Biol., 2010, 5, 489–497.
31 T. M. Gloster and D. J. Vocadlo, Nat. Chem. Biol., 2012, 8, 683–694.
Chair in Structural Biology. DJV is a Canada Research Chair in 32 B. Moya, A. Dotsch, C. Juan, J. Blazquez, L. Zamorano, S. Haussler
and A. Oliver, PLoS Pathog., 2009, 5, e1000353.
33 L. Zamorano, T. M. Reeve, L. Deng, C. Juan, B. Moya, G. Cabot,
Chemical Glycobiology and an EWR Steacie Memorial Fellow. We
also thank V. Larmour, Z. Madden, and the staff of the Canadian
D. J. Vocadlo, B. L. Mark and A. Oliver, Antimicrob. Agents Chemother.,
Light Source (CLS) beamline 08ID-1 for assistance. Coordinates and
2010, 54, 3557–3563.
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 10983--10985 10985