Synthesis of LuxS inhibitors targeting bacterial cell-cell communication

Article abstract of DOI:10.1021/ol049182i

(Chemical Equation Presented) Quorum sensing is a process by which bacteria sense cell density. This cell-cell communication process is mediated by autoinducers. A cross-species messenger, autoinducer-2 (Al-2) is produced from S-ribosyl-L-homocysteine by the LuxS enzyme. A proposed mechanism for LuxS is an aldose-ketose isomerization of S-ribosylhomocysteine followed by a β-elimination. We report here the synthesis of two substrate analogues, S-anhydroribosyl-L-homocysteine and S-homoribosyl-L-cysteine, which prevent the initial and final step of the mechanism, respectively.

Full text of DOI:10.1021/ol049182i

ORGANIC
LETTERS
2004
Vol. 6, No. 18
3043-3046
Synthesis of LuxS Inhibitors Targeting
Bacterial Cell−Cell Communication
Joshua F. Alfaro,† Tiao Zhang,† DonRaphael P. Wynn,†,‡,§
,†,‡,§,|
Erin L. Karschner,†,|, and Zhaohui Sunny Zhou*
Department of Chemistry, School of Molecular Biosciences, Center for Integrated
Biotechnology, Graduate Program in Pharmacology and Toxicology,
Washington State UniVersity, Pullman, Washington 99164-4630
sunnyz@wsu.edu
Received May 4, 2004
ABSTRACT
Quorum sensing is a process by which bacteria sense cell density. This cell−cell communication process is mediated by autoinducers. A
cross-species messenger, autoinducer-2 (AI-2) is produced from S-ribosyl-L-homocysteine by the LuxS enzyme. A proposed mechanism for
LuxS is an aldose-ketose isomerization of S-ribosylhomocysteine followed by a â-elimination. We report here the synthesis of two substrate
analogues, S-anhydroribosyl-L-homocysteine and S-homoribosyl-L-cysteine, which prevent the initial and final step of the mechanism, respectively.
Although bacteria are monocellular organisms, they com-
municate actively within their own species and across species
boundaries, allowing them to form highly diverse communi-
ties, such as biofilms. One of the cell-cell communication
processes by which they sense cell density is termed quorum
sensing.1-3 Winas and Bassler called quorum sensing “Mob
Psychology”. They wrote, “the (external) signal allows
bacteria to sense when they have achieved a ‘quorum’. The
quorum contains a sufficient number of bacteria to carry out
processes that necessitates the cooperation of a large number
of cells in order to be effective.”2 Quorum sensing is found
in a broad spectrum of bacterial species and regulates toxin
production, biofilm formation, sporulation, and virulence
gene expression, for example.1 Quorum sensing is mediated
by autoinducers (AIs), small signaling molecules generated
by bacteria. In general, Gram-negative bacteria produce
N-acyl-homoserine lactones as autoinducers, whereas Gram-
positive species generate peptides as signals.1-3 Unlike other
autoinducers that are species-specific, autoinducer-2 (AI-2,
a furanosyl borate diester, see Scheme 1) mediates cross-
species communication.3-5 For instance, Pseudomonas aerug-
inosa does not produce AI-2. However, Surette and co-
workers found that AI-2 induces several virulence genes in
P. aeruginosa; moreover, AI-2 was generated in the oropha-
ryngeal flora in sputum samples obtained from cystic fibrosis
patients.6 These findings underscore the impact of environ-
ment and microbiota shift in bacterial virulence.
† Department of Chemistry.
‡ School of Molecular Biosciences.
§ Center for Integrated Biotechnology.
| Graduate Program in Pharmacology and Toxicology.
Departments of Chemistry and Biology, Lycoming College, Williams-
port, PA 17701.
(4) Chen, X.; Schauder, S.; Potier, H.; Dorsselaer, A. V.; Pelczer, I.;
Bassler, B. L.; Hughson, F. M. Nature 2002, 415, 545.
(5) Winzer, K.; Hardie, K. R.; Williams, P. AdV. Appl. Microbiol. 2003,
53, 291.
(1) Fuqua, C.; Parsek, M. R.; Greenberg, E. P. Annu. ReV. Genet. 2001,
35, 439.
(2) Winans, S. C.; Bassler, B. L. J. Bacteriol. 2002, 184, 873.
(3) Xavier, K. B.; Bassler, B. L. Curr. Opin. Microbiol. 2003, 6, 191.
(6) Duan, K.; Dammel, C.; Stein, J.; Rabin, H.; Surette, M. G. Mol.
Microbiol. 2003, 50, 1477.
10.1021/ol049182i CCC: $27.50 © 2004 American Chemical Society
Published on Web 08/04/2004

Scheme 1. Formation of Autoinducer 2 (AI-2)
Scheme 2. Proposed Mechanism for LuxS
boxed).15,18-20 In the initial steps (1a to 1c), an aldose-ketose
isomerization generates a ketone at the C3 position on the
carbohydrate moiety. Similar to other aldose-ketose isomer-
ases, LuxS also contains a divalent metal ion in the active
site.21-23 However, aldose-ketose isomerization through two
bonds, as proposed in the LuxS reaction, is unprecedented
for enzyme-catalyzed transformations.24,25 Thus, LuxS sports
an intriguing mechanism on its own. In the final step of the
proposed mechanism (2 in Scheme 2), a base in LuxS
abstracts the C4 proton and eliminates the homocysteinyl
thiol, and then the enol intermediate formed spontaneously
rearranges into the DPD product. Hence, from a mechanistic
point of view, LuxS is analogous to S-adenosylhomocysteine
hydrolase.26,27 As proposed, both generate a C3-ketone
intermediate before cleaving the carbon-sulfur bond via
â-elimination. On the other hand, the two enzymes differ in
the way ketone intermediates are generated. The AdoHcy
hydrolase produces the C3 ketone via nicotinamide adenine
dinucleotide (NAD)-dependent redox chemistry,26,27 whereas
LuxS may produce the putative C3 ketone intermediate via
an aldo-ketose isomerization. In essence LuxS possesses the
function of both an aldo-ketose isomerase and a lysase.
The ubiquitous AI-2 formation pathway is found in about
half of bacterial species (see Scheme 1).3,5 It begins with
S-adenosyl-L-homocysteine (AdoHcy), a common product
of S-adenosyl-L-methionine (AdoMet)-dependent methyl-
ation, a large family of transformations present in all
organisms.7 AdoHcy is hydrolyzed to S-ribosyl-L-homocys-
teine (SRH) and adenine by S-adenosyl-L-homocysteine/
5′-methylthioadenosine nucleosidase (SAHN or MTAN,
EC 3.2.2.9).8-10 Then the enzyme LuxS cleaves S-ribosyl-
homocysteine to form L-homocysteine (Hcy) and 4,5-dihy-
droxy-2,3-pentanedione (DPD); the latter is the precursor of
autoinducer-2.11-15 LuxS is likely to be the enzyme ribosyl-
homocysteinase (EC 3.2.1.148; formerly S-ribosyl-L-homo-
cysteine hydrolase, EC 3.3.1.3), which was first described
by Duerre and co-workers in the 1960s.16,17
As shown in Scheme 2, Pei’s group and our group have
proposed a mechanism for LuxS (the overall reaction is
(7) S-Adenosylmethionine-Dependent Methyltransferases: Structures and
Functions; Cheng, X., Blumenthal, R., Eds.; World Scientific: Singapore,
River Edge, NJ, 1999.
(8) Cornell, K. A.; Swarts, W. E.; Barry, R. D.; Riscoe, M. K. Biochem.
Biophys. Res. Commun. 1996, 228, 724.
(9) Lee, J. E.; Cornell, K. A.; Riscoe, M. K.; Howell, P. L. Acta
Crystallogr., Sect. D 2001, 57, 150.
(10) Lee, J. E.; Cornell, K. A.; Riscoe, M. K.; Howell, P. L. Structure
2001, 9, 941.
(11) Schauder, S.; Shokat, K.; Surette, M. G.; Bassler, B. L. Mol.
Microbiol. 2001, 41, 463.
(12) Surette, M. G.; Miller, M. B.; Bassler, B. L. Proc. Natl. Acad. Sci.
U.S.A. 1999, 96, 1639.
(13) Taga, M. E.; Semmelhack, J. L.; Bassler, B. L. Mol. Microbiol.
2001, 42, 777.
(14) Winzer, K.; Hardie, K. R.; Burgess, N.; Doherty, N.; Kirke, D.;
Holden, M. T.; Linforth, R.; Cornell, K. A.; Taylor, A. J.; Hill, P. J.;
Williams, P. Microbiology 2002, 148, 909.
(15) Zhao, G.; Wan, W.; Mansouri, S.; Alfaro, J. F.; Bassler, B. L.;
Cornell, K. A.; Zhou, Z. S. Bioorg. Med. Chem. Lett. 2003, 13, 3897.
(16) Duerre, J. A.; Miller, C. H. J. Bacteriol. 1966, 91, 1210.
(17) Duerre, J. A.; Baker, D. J.; Salisbury, L. Fed. Proc. 1971, 30, 88.
(18) Zhu, J.; Dizin, E.; Hu, X.; Wavreille, A. S.; Park, J.; Pei, D.
Biochemistry 2003, 42, 4717.
(19) Ritter, K. A. Dissertation, Washington State University, 2003.
(20) Zhu, J.; Hu, X.; Dizin, E.; Pei, D. J. Am. Chem. Soc. 2003, 125,
13379.
(21) Lewis, H. A.; Furlong, E. B.; Laubert, B.; Eroshkina, G. A.;
Batiyenko, Y.; Adams, J. M.; Bergseid, M. G.; Marsh, C. D.; Peat, T. S.;
Sanderson, W. E.; Sauder, J. M.; Buchanan, S. G. Structure (Camb) 2001,
9, 527.
(22) Ruzheinikov, S. N.; Das, S. K.; Sedelnikova, S. E.; Hartley, A.;
Foster, S. J.; Horsburgh, M. J.; Cox, A. G.; McCleod, C. W.; Mekhalfia,
A.; Blackburn, G. M.; Rice, D. W.; Baker, P. J. J. Mol. Biol. 2001, 313,
111.
(23) Hilgers, M. T.; Ludwig, M. L. Proc. Natl. Acad. Sci. U.S.A. 2001,
98, 11169.
(24) Schoss, J. V.; Hixon, M. S. In ComprehensiVe Biological Catalysis;
Sinnott, M., Ed.; Academic Press: San Diego, 1998; p 43.
(25) Watt, C. I. F. In ComprehensiVe Biological Catalysis; Sinnott, M.,
Ed.; Academic Press: San Diego, 1998; p 253.
(26) Takata, Y.; Yamada, T.; Huang, Y.; Komoto, J.; Gomi, T.; Ogawa,
H.; Fujioka, M.; Takusagawa, F. J. Biol. Chem. 2002, 277, 22670.
(27) Turner, M. A.; Yuan, C.-S.; Borchardt, R. T.; Hershfield, M. S.;
Smith, G. D.; Howell, P. L. Nat. Struct. Biol. 1998, 5, 369.
3044
Org. Lett., Vol. 6, No. 18, 2004

the only substrate-enzyme complex structure was observed
with an inactive form of the enzyme, in which a conserved
cysteine residue was oxidized (Cys84 in the B. subtilis
protein).22,23 Because compound 1 is capable of forming a
stable complex with the active form of the LuxS enzyme,
the structural information gleaned using this substrate
analogue will shed light on the interaction between the
substrate and the active enzyme and assist structure-based
inhibitor design as well.
The second substrate analogue (compound 2 in Figure 1)
is S-homoribosyl-L-cysteine (S-(5,6-dideoxy-D-ribo-hexo-
furanos-6-yl)-L-cysteine). As illustrated in Scheme 4, oxida-
Figure 1. LuxS substrate and two substrate analogues.
Scheme 4. Synthesis of S-Homoribosyl-L-cysteine (2)
Absent in humans, the LuxS enzyme is an attractive target
for novel therapeutic agent development for bacterial infec-
tion.3,5 Prior to this report, however, there was no report of
a specific LuxS inhibitor. Herein we report the synthesis of
two LuxS substrate analogues that function as inhibitors and
mechanistic probes.
The first substrate analogue (compound 1 in Figure 1) is
S-anhydroribosyl-L-homocysteine (S-[1,4-anhydro-5-deoxy-
D-ribitol-5-yl]-homocysteine). As depicted in Scheme 3,
Scheme 3. Synthesis of S-Anhydroribosyl-L-homocysteine (1)
tion reduction condensation followed by standard deprotec-
tion procedures afforded the target molecule.
In compound 2, the C5-C6 carbon-carbon bond replaces
the C5 carbon-sulfur bond of the S-ribosylhomocysteine
substrate, effectively making carbon-sulfur bond cleavage
impossible (Scheme 5). On the other hand, the ribose moiety
Scheme 5. Accumulation of Ketone Intermediate by
Preventing â-Elimination
sulfide 5 was constructed via oxidation reduction condensa-
tion between the fully protected homocystine and anhydrori-
bose;28,29 both were prepared by literature procedures.
Standard deprotection methods afforded the desired product.
We rationalized that replacement of the hemiacetal in the
S-ribosylhomocysteine substrate by an ether in compound 1
would prevent initial aldo-ketose isomerization.15,18-20 On
the other hand, S-anhydroribosyl-homocysteine still possesses
the 2,3-diol for ligation to the active site metal ion.22 Thus,
compound 1 is likely to bind to LuxS in a similar fashion as
the substrate and could serve as a LuxS inhibitor. Up to now,
and the amino acid moiety of the substrate and compound 2
are connected by the same number of C-C and C-S bonds.
As a result, this substrate analogue is expected to be able to
bind to LuxS in a productive orientation, i.e., the compound
should still be able to undergo initial aldo-ketose isomer-
izations to form a ketone at the C3 position (see Scheme 5).
At this point, however, the C3 ketone cannot undergo further
elimination reaction, which might allow direct observation
of the keto intermediate via X-ray crystallography or by
trapping with chemical reagents. This would provide direct
(28) Mukaiyama, T.; Takei, H. Top. Phosphorus Chem. 1976, 8, 587.
(29) Serafinowski, P.; Dorland, E.; Harrap, K. R.; Balzarini, J.; De Clercq,
E. J. Med. Chem. 1992, 35, 4576.
Org. Lett., Vol. 6, No. 18, 2004
3045

Using our previously reported LuxS assay,15,19 the pre-
liminary studies show that LuxS does not cleave the C-S
bond of compounds 1 and 2; see Figure 2 (see Supporting
Information for compound 1 data). Moreover, both com-
pounds (1 and 2) inhibit the LuxS enzyme, see Figure 3 (see
Supporting Information for compound 2 data).
With the ready availability of the first generation of
specific LuxS inhibitors, we are now poised to investigate
the effects of LuxS inhibition in both in vitro and in vivo
studies. The results of our continuing work in this area will
be reported in due course.
Figure 2. Time courses of reaction between S-ribosyl-homocys-
teine (100 µM) and LuxS (0.34 µM, b); compound 2 (100 µM)
and LuxS (0.34 µM, +; 3.4 µM, (); compound 2 (300 µM) and
LuxS (3.4 µM, 4); LuxS alone (0.34 µM, O); and S-ribosyl-
homocysteine alone (100 µM, 0).
Acknowledgment. We are indebted to Drs. Martha L.
Ludwig and Mark T. Hilgers at the University of Michigan
for providing the LuxS expression strain and their support
and Drs. Kenneth A. Cornell and Michael K. Riscoe at the
Portland VA Medical Center for providing the AdoHcy
nucleosidase expression strain. This work was supported by
the National Institute of Allergy and Infectious Diseases
(NIAID/NIH, 1R01AI058146 to Z.S.Z.) and the Herman
Frasch Foundation (541-HF02 to Z.S.Z.). The WSU NMR
Center equipment was supported by NIH grants RR0631401
and RR12948, NSF grants CHE-9115282 and DBI-9604689,
and the Murdock Charitable Trust. J.F.A. was supported in
part by the M.S. to Ph.D. transition award provided by the
WSU Graduate School. D.P.W. was supported by a summer
research fellowship from the Center for Integrated Biotech-
nology at Washington State University. E.L.K. was supported
by the Summer Undergraduate Research Fellowship (SURF)
program in Pharmacology and Toxicology at Washington
State University. We thank Drs. Rob Ronald and Pat Meier
for helpful comments on the manuscript and Dr. Greg Helms
for assistance regarding NMR.
support for the proposed involvement of the keto intermediate
in LuxS catalysis and, again, structural information for
rational inhibitor design.
Supporting Information Available: Experimental pro-
cedures for the synthesis of compounds 1 and 2. This material
is available free of charge via the Internet at http://pubs.acs.org.
Figure 3. Reaction courses of S-ribosyl-homocysteine (100 µM)
and LuxS (0.34 µM) in the presence of compound 1 (0 µM, b;
100 µM, 0; 300 µM, ); and 1000 µM, (); and LuxS alone (0.34
µM, O).
OL049182I
3046
Org. Lett., Vol. 6, No. 18, 2004