Identification of active-site inhibitors of MurG using a generalizable, high-throughput glycosyltransferase screen

Article abstract of DOI:10.1021/ja036494s

MurG is a glycosyltransferase involved in the biosynthesis of bacterial peptidoglycan. It is a potentially important antibiotic target, but no inhibitors of the enzyme have been reported. In general, inhibitors of glycosyltransferases have been difficult

Full text of DOI:10.1021/ja036494s

Published on Web 08/22/2003
Identification of Active-Site Inhibitors of MurG Using a Generalizable,
High-Throughput Glycosyltransferase Screen
Jeremiah S. Helm, Yanan Hu, Lan Chen, Ben Gross, and Suzanne Walker*
Department of Chemistry, Princeton UniVersity, Princeton, New Jersey 08544
Received June 3, 2003; E-mail: swalker@princeton.edu
Peptidoglycan is a cross-linked carbohydrate polymer that forms
layers around bacterial cell membranes.1 One of its primary
functions is to protect bacterial cells from lysis due to fluctuations
in internal osmotic pressure. The machinery for peptidoglycan
biosynthesis is highly conserved in both Gram-negative and Gram-
positive bacteria, and each of the enzymes involved in the pathway
Figure 1. Reaction catalyzed by MurG.
is a potential target for antibiotic chemotherapy.2 In recent years,
inhibitors have been identified for all of the peptidoglycan-
synthesizing enzymes except MurG,3 the enzyme that catalyzes the
final intracellular step in the biosynthetic pathway (Figure 1).4 MurG
is a member of a superfamily of nucleotide-sugar glycosyltrans-
ferases (NDP-Gtases), enzymes that catalyze the transfer of sugars
from nucleotides such as UDP/TDP or GDP to a variety of
acceptors.5 Like kinases, nucleotide-glycosyltransferases are ubiq-
uitous and play roles in a wide range of biological processes, some
of which are pathogenic. Despite considerable effort, it has generally
been difficult to design good inhibitors of Gtases.6 Furthermore,
the most potent inhibitors typically contain negatively charged
diphosphates or mimics thereof, and thus have limited potential as
leads.
High-throughput screening (HTS) has been used successfully to
Figure 2. (a) Electrostatic surface representation of the MurG binding cleft
showing the exposed C2 N-acyl group of UDP-GlcNAc.7 The uracil ring is
buried. (b) Structure of the labeled UDP-GlcNAc probe.
identify neutral inhibitors of protein kinases and phosphatases that
compete with the negatively charged substrates of these enzymes.8
Therefore, we thought it might be possible, by screening large
numbers of compounds, to identify inhibitors containing structural
elements that mimic the functions of the diphosphoryl group of
the nucleotide-sugar donor used by MurG. These functions include
positioning the uridine and hexose substituents and contributing
favorably to binding. Here, we describe a fluorescence-based
substrate displacement assay that was used to screen about 50,000
compounds against MurG.9 The assay, which has allowed us to
identify a promising family of MurG inhibitors, can be readily
adapted to screen other glycosyltransferases for potential inhibitors.
Our goal was to identify compounds that compete with UDP-
GlcNAc for binding to the active site of MurG in the hope that
some of these compounds would contain structural elements that
mimic the functions of the diphosphate. We decided that the best
way to enrich screening hits for competitors of UDP-GlcNAc would
be to use a high-throughput assay based on displacement of the
hexose donor from the active site. The X-ray structure of a co-
complex of MurG containing UDP-GlcNAc shows that the C2
N-acetyl group on the donor is solvent exposed and the protein
makes no contacts to the methyl group (Figure 2a).7 Therefore, we
anticipated that donor analogues containing N-acyl modifications
would bind to MurG. Accordingly, we synthesized the fluoresce-
inated UDP-GlcNAc analogue 1 (Figure 2b) and evaluated it for
use in a displacement assay. The anisotropy of 1 increased
significantly in the presence of MurG. From the change in
anisotropy as a function of MurG concentration, we calculated a
dissociation constant of 1.7 ( 0.2 µM for 1. Adding unlabeled UDP-
GlcNAc, UDP, or UMP to a preequilibrated mixture of MurG and
1 caused a decrease in anisotropy, showing that all three of these
molecules displace 1 from the active site, albeit at different
concentrations. Dissociation constants for the three ligands were
calculated from the concentration dependence of the anisotropy
change.10 The KD of UDP-GlcNAc was found to be identical to
that of 1, implying that the fluorophore does not affect binding.
The KD of UDP (2.3 µM) was found to be much better than that of
UMP (89 µM), consistent with the IC50 values measured previ-
ously.11 These experiments suggested that it would be possible to
distinguish weaker binders from better binders on the basis of
changes in anisotropy at fixed concentrations of 1, MurG, and the
putative binder.
We developed a miniaturized fluorescence polarization assay and
used it to screen 48,877 compounds in duplicate over 5 days at the
Institute for Chemistry and Cell Biology, a collaborative screening
facility located at Harvard Medical School. Compounds were
screened in 384-well plates. Each plate contained two control wells,
one with MurG and 1 alone and the other with MurG, 1, and 25
µM UDP. Under the assay conditions, the well containing UDP
gave a polarization reading that was approximately 50% that of
the wells containing only MurG and 1. Test compounds were added
in 100 nL of DMSO to the sample wells to give a final concentration
of 25 µg/mL in a final volume of 20 µL. Because our aim was to
identify ligands that approached or exceeded UDP in binding to
MurG, we scored as positive only those wells in which the
fluorescence signal reproducibly dropped by more than 50%. There
were 277 compounds that scored positive by this criterion, for a
hit rate of 0.6%. The true hit rate is undoubtedly lower than this
because a significant number of the hit compounds were fluorescent.
9
11168
J. AM. CHEM. SOC. 2003, 125, 11168-11169
10.1021/ja036494s CCC: $25.00 © 2003 American Chemical Society

C O M M U N I C A T I O N S
The displacement assay used here for high-throughput screening
is simple to implement and can be readily adapted to screen any
glycosyltransferase in which at least one modifiable group on the
nucleotide-sugar is solvent exposed. Crystal structures of existing
glycosyltransferases can serve as starting points for the design of
suitable fluorescent substrate analogues, and structural similarities
between glycosyltransferases imply that fluorescent analogues that
work for one GTase will also work for related GTases.5 It may be
possible to identify families of scaffolds that mimic diphosphates
in different conformations. Such a set of scaffolds would be
invaluable for the diversity-oriented synthesis of libraries to be
screened for glycosyltransferase inhibition.
Figure 3. Representative compound from the family of MurG inhibitors
with the common core highlighted.
Acknowledgment. This work was supported by NIH grant
A144854. We thank Dr. Caroline Shamu, James Follen, Stewart
Rudnicki, Katrina Schulberg, Dara Greenhouse, and the staff at
the ICCB/Harvard Medical School for their assistance.
Supporting Information Available: Synthetic scheme and char-
acterization of 1; anisotropy curves; experimental details for secondary
screening; assay conditions and inhibition pattern for compound 2
(PDF). This material is available free of charge via the Internet at http://
pubs.acs.org.
Figure 4. Overlay of 2 (red) with UDP-GlcNAc (blue) bound to the MurG
crystal structure. Some residues have been cut away to expose the uracil
ring.
References
(1) Seltmann, G.; Holst, O. The Bacterial Cell Wall; Springer: New York,
2002.
We selected 44 compounds, representing several different
structural classes, for secondary screening using a kinetic assay in
which the production of radiolabeled product was measured.12
Compounds were tested in duplicate at a concentration of ∼5 µM.
At this concentration and under the same assay conditions, UDP
inhibited the enzymatic reaction by 50%. As before, we wanted
compounds that were similar or better inhibitors than UDP, and so
we selected only those that reproducibly inhibited the enzyme by
more than 50% under the assay conditions. Eleven of the 44
compounds met this criterion. Seven of these 11 compounds have
a five-membered, nitrogen-containing heterocyclic core with an
alkyl or aryl substituent at N-1 and an arylidene substituent at the
3 position. A representative of the family, and the most potent of
the seven related inhibitors, is shown in Figure 3. Compound 2 is
a competitive inhibitor of MurG with respect to the UDP-GlcNAc
substrate (Supporting Information).
The high percentage of inhibitors with a similar core is striking
and suggests that these compounds share a common binding mode.
Manual docking of the inhibitors into the UDP-GlcNAc binding
pocket of MurG reveals that the compounds are best accommodated
when the five-membered ring is located in the vicinity of the
diphosphate binding site with the N-1 substituent oriented toward
the GlcNAc binding site and the arylidene substituent oriented
toward the uridine binding site (Figure 4).13 Thus, by using a NDP-
sugar displacement assay, we have identified a family of MurG
inhibitors with a neutral core that may mimic the diphosphate
moiety of UDP-GlcNAc with respect to the display of substituent
groups. We are currently trying to obtain crystals of MurG with
some of these inhibitors to evaluate the proposed binding mode
and to provide information to guide the design of better inhibitors.
(2) Walsh, C. T. Antibiotics: Actions, Origins, Resistance; ASM Press:
Washington, DC, 2003.
(3) (a) Wong, K. K.; Pompliano, D. L. AdV. Exp. Med. Biol. 1998, 456, 197-
217. (b) El Zoeiby, A.; Sanschagrin, F.; Levesque, R. C. Mol. Microbiol.
2003, 47, 1-12.
(4) (a) Salmond, G. P.; Lutkenhaus, J. F.; Donachie, W. D. J. Bacteriol. 1980,
144, 438-440. (b) Ikeda, M.; Wachi, M.; Jung, H. K.; Ishino, F.;
Matsuhashi, M. Nucleic Acids Res. 1990, 18, 4014. (c) Mengin-Lecreulx,
D.; Texier, L.; Rousseau, M.; van Heijenoort, J. J. Bacteriol. 1991, 173,
4625-4636. (d) Bupp, K.; van Heijenoort, J. J. Bacteriol. 1993, 175,
1841-1843. (e) Men, H.; Park, P.; Ge, M.; Walker, S. J. Am. Chem. Soc.
1998, 120, 2484-2485. (f) Chen, L.; Men, H.; Ha, S.; Ye, X.-Y.; Brunner,
L.; Hu, Y.; Walker, S. Biochemistry 2002, 41, 6824-6833. (g) Ha, S.;
Walker, D.; Shi, Y.; Walker, S. Protein Sci. 2000, 9, 1045-1052.
(5) Hu, Y.; Walker, S. Chem. Biol. 2002, 9, 1287-1296.
(6) (a) Wang, R.; Steensma, D. H.; Takaoka, Y.; J.W., Y.; Kajimoto, T.; Wong,
C.-H. Bioorg. Med. Chem. 1997, 5, 661-672. (b) Qian, X.; Palcic, M.
M. Carbohydr. Chem. Biol. 2000, 3, 293-312. (c) Saotome, C.; Wong,
C.-H.; Kanie, O. Chem. Biol. 2001, 8, 1061-1070. (d) Compain, P.;
Martin, O. R. Bioorg. Med. Chem. 2001, 9, 3077-3092. (e) Compain,
P.; Martin, O. R. Curr. Top. Med. Chem. 2003, 3, 541-560.
(7) Hu, Y.; Chen, L.; Ha, S.; Gross, B.; Falcone, B.; Walker, D.; Mokhtarza-
deh, M.; Walker, S. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 845-849.
(8) (a) Bridges, A. J. Chem. ReV. 2001, 101, 2541-2572. (b) Zhang, Z. Y.
Annu. ReV. Pharmacol. Toxicol. 2002, 42, 209-234.
(9) Hertzberg, R. P.; Pope, A. J. Curr. Opin. Chem. Biol. 2000, 4, 445-451.
(10) Cho, J.; Hamasaki, K.; Rando, R. R. Biochemistry 1998, 37, 4985-4992.
(11) Ha, S.; Chang, E.; Lo, M.-C.; Men, H.; Park, P.; Ge, M.; Walker, S. J.
Am. Chem. Soc. 1999, 121, 8415-8426.
(12) Helm, J. S.; Chen, L.; Walker, S. J. Am. Chem. Soc. 2002, 124, 13970-
13971.
(13) Similar functionalized cores have been reported for inhibitors of other
enzymes that utilize UDP-sugars. See: (a) Andres, C. J.; Bronson, J. J.;
D’Andrea, S. V.; Deshpande, M. S.; Falk, P. J.; Grant-Young, K. A.;
Harte, W. E.; Ho, H. T.; Misco, P. F.; Robertson, J. G.; Stock, D.; Sun,
Y.; Walsh, A. W. Bioorg. Med. Chem. Lett. 2000, 10, 715-717. (b) Ma,
Y.; Stern, R. J.; Scherman, M. S.; Vissa, V. D.; Yan, W.; Jones, V. C.;
Zhang, F.; Franzblau, S. G.; Lewis, W. H.; McNeil, M. R. Antimicrob.
Agents Chemother. 2001, 45, 1407-1416. (c) Sim, M. M.; Ng, S. B.;
Buss, A. D.; Crasta, S. C.; Goh, K. L.; Lee, S. K. Bioorg. Med. Chem.
Lett. 2002, 12, 697-699.
JA036494S
9
J. AM. CHEM. SOC. VOL. 125, NO. 37, 2003 11169