Angewandte

Chemie

DOI: 10.1002/anie.201307975
Antimicrobial Surfaces

Bacteria-Triggered Release of Antimicrobial Agents**

Vitaly V. Komnatnyy, Wen-Chi Chiang, Tim Tolker-Nielsen, Michael Givskov, and
Thomas E. Nielsen*

Dedicated to Professor Morten Meldal on the occasion of his 60th birthday

Abstract: Medical devices employed in healthcare practice are for functionalization of material surfaces that prevent biofilm
often susceptible to microbial contamination. Pathogenic formation would clearly have significant commercial poten-
bacteria may attach themselves to device surfaces of catheters tial and socioeconomic benefits. Along these lines, various
or implants by formation of chemically complex biofilms, methods, such as impregnating the devise material, for
which may be the direct cause of device failure. Extracellular example, polyurethane walls of central-venous catheter,
bacterial lipases are particularly abundant at sites of infection. with bactericidal components, along with different coating
Herein it is shown how active or proactive compounds attached strategies, have been developed.[4]
to polymeric surfaces using lipase-sensitive linkages, such as Here, we present a new approach for the bacteria-
fatty acid esters or anhydrides, may be released in response to triggered release of antimicrobial compounds (Scheme 1)
infection. Proof-of-concept of the responsive material is from poly(ethylene glycol) (PEG) materials, thereby paving
demonstrated by the bacteria-triggered release of antibiotics the way for a new generation of bacteria-repellent materials.
to control bacterial populations and signaling molecules to Extracellular lipases are particularly abundant at sites of
modulate quorum sensing. The self-regulating system provides bacterial infection, as indicated by high amounts of anti-lipase
the basis for the development of device-relevant
polymeric materials, which only release antibi-
otics in dependency of the titer of bacteria
surrounding the medical device.

M edical devices employed in healthcare prac-

tice are often susceptible to microbial contami-
nation.[1] Pathogenic bacteria may attach them-
selves to device surfaces of catheters or implants
by formation of chemically complex biofilms,
which may be the direct cause of device failure. Scheme 1. The concept of bacteria-trigged enzymatic release of antibiotics from
Treatment of device-related infections is often chemically modified polymers.
difficult as pathogens exhibit a high degree of
antibiotic resistance in the biofilm mode-of-
life.[2] For the most severe cases, surgery and replacement of in anti-sera obtained from cystic fibrosis patients suffering
the indwelling device may be the only option for patient from pseudomonas aeruginosa infection.[5] We therefore
[3]
survival. In this context, the development of new methods envisioned that active or proactive compounds attached to
polymeric surfaces using lipase-sensitive linkages, such as
fatty acid esters or anhydrides, could be released in response
[*] Dr. V. V. Komnatnyy, Prof. Dr. T. E. Nielsen
Department of Chemistry, Technical University of Denmark
to infection. We communicate our recent investigations on
Kgs. Lyngby, 2800 (Denmark) such bacteria-responsive materials and report a novel lipase-
E-mail: [email protected] triggered release system for the control of bacterial popula-
Dr. W.-C. Chiang, Prof. Dr. T. Tolker-Nielsen, Prof. Dr. M. Givskov tions and modulation of quorum sensing.
Department of International Health It is commonly believed that once biofilms are formed,
Immunology and Microbiology, University of Copenhagen bacterial cell-to-cell signaling, frequently referred to as
Copenhagen, 2200 (Denmark) quorum sensing (QS), plays an essential role for bacteria in
Prof. Dr. M. Givskov, Prof. Dr. T. E. Nielsen their evasion of the host immune system and resistance to
Singapore Centre on Environmental Life Sciences Engineering antibiotic treatment.[6] Although host phagocytic leukocytes
Nanyang Technological University
(PMNs) and administered antibiotics can normally clear
637551 Singapore (Singapore)
acute infections that originate from planktonic bacteria, those
[**] DSF Center for Antimicrobial Research (Danish Council for Strategic
Research) is gratefully acknowledged for financial support. We measures fail to eradicate bacteria present in biofilms because
thank Dr. Susanne Wilhelm, Heinrich-Heine-University Dsseldorf, of protection offered by the extracellular biofilm matrix.
for providing the P. aeruginosa lipA lipC estA mutant. Given the importance of quorum sensing and cell–cell
Supporting information for this article is available on the WWW signaling for infectious processes, release systems that can
under http://dx.doi.org/10.1002/anie.201307975. be triggered under carefully controlled conditions could

Angew. Chem. Int. Ed. 2014, 53, 439 –441  2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 439
 .
Angewandte
Communications

Scheme 2. Solid-phase synthesis of homoserine-containing ester 4 and

lipase-triggered release of C4-HSL 5: 1) Fmoc-O-trityl-homoserine, 1-
(mesitylene-2-sulfonyl)-3-nitro-1,2,4-triazole (MSNT), N-methyl-imida-
zole, DMF, 1 h, RT; 2) 20 % piperidine (DMF), RT; 3) butanoyl chlo-
ride, NEt3, CH2Cl2, 1 h, RT; 4) 5 % TFA in CH2Cl2, 1 h, RT; 5) octanoyl
Figure 1. C4-HSL-dependant activation of the ahyR/ahyI-gfp reporter
chloride, triethylamine, DMAP, CH2Cl2, 1 h, RT; 6) lipase from Pseudo-
system in E. coli (OD450 = optical density at 450 nm).
monas fluorescens (Fmoc = 9-fluorenylmethoxycarbonyl, DMF = N,N-
dimethylformamide, TFA = trifluoroacetic acid, DMAP = 4-dimethylami-
nopyridine, and HMBA = 4-hydroxymethylbenzoic acid).
Having developed a surface that releases functional C4-
HSL in response to defined lipase-activity, we shifted our
become biomedically relevant to treat and diagnose popula- focus to a system capable of also releasing antimicrobial
tions of Gram-negative bacteria, such as P. aeruginosa, drugs. Ciprofloxacin was chosen because of its function as
through QS modulation.[7] Our attention turned to the acyl- a broad-spectrum antibiotic. This compound contains a car-
homoserine lactone (AHL) molecules produced by P. aeru- boxylic acid moiety which can be used as an anchoring
ginosa, N-3-oxo-dodecanoyl-l-homoserine lactone (3O-C12- element for chemical synthesis of lipase-labile bonds. Along
HSL) and N-butyryl-l-homoserine lactone (C4-HSL), which these lines, mixed anhydride construct 8 was assembled on
directly or indirectly control the expression of multiple ChemMatrix resin according to the synthetic sequence shown
virulence factors, secondary metabolites, swarming motility, at the Scheme 3. This construct is completely stable under
and biofilm development.
The homoserine-containing construct 4 (Scheme 2) was
assembled on HMBA-linked ChemMatrix polymer,[8] which
conveniently serves as a representative model system for
PEG-based materials (HMBA = 4-hydroxymethylbenzoic
acid. Like related materials, this polymer allows diffusion of
biological macromolecules, such as enzymes and substrates.[9]
We envisioned how lipase-treatment of 4 upon ester
hydrolysis would liberate a free hydroxy group and undergo
a spontaneous cyclization, possibly favored by the acidic
interior of the lipase, and thereby release C4-HSL. We chose Scheme 3. Solid-phase synthesis of mixed anhydride 8 and bacteria-
the C4-HSL signal molecule because this allows easy assess- triggered release of ciprofloxacin: 1) azelaic anhydride, N-methyl
ment of the effect of the release system on bacteria as imidazole, CH2Cl2, 2 h; 2) Boc-ciprofloxacin, bis(trichloromethyl)car-
described below. In addition, the system can easily be bonate, triethylamine, CH2Cl2, 2 h; 3) TMSOTf, CH2Cl2, 2 h; 4) Pseudo-
modified to release quorum quenching compounds. monas fluorescens lipase (Boc = tert-butoxycarbonyl and TMSOTf = tri-
methylsilyltrifluoromethanesulfonate).
To test the approach we employed E. coli MH205 as a QS
monitor strain. This monitor contains an ahyR/ahyI-gfp
reporter system, which responds readily to the presence of
extracellular C4-HSL, with the E. coli strain serving as physiological pH conditions, but upon shifting to basic media
a lipase-negative background. We prepared a set of experi- both ciprofloxacin and the azelaic acid spacer (itself an
ments where beads 4 (resin loading of about 0.3 mmol g 1) antimicrobial compound) were released from the support
were suspended in bacterial growth medium in concentrations (Scheme 3). We envisioned that the presence of extracellular
corresponding roughly to 100 mm C4-HSL at full liberation. As bacterial lipases would catalyze the hydrolysis of the mixed
controls, a 2.5 mm solution of C4-HSL-supplemented growth anhydride bond and thus liberate the antibiotic.
medium as well as an unsupplemented growth medium were As to this end, we investigated if the antibiotic-coated
included. Rewardingly, beads 4 and the presence of Pseudo- beads 8 could actually kill strains of P. aeruginosa. This
monas fluorescens lipase were indeed found to induce biofilm-forming opportunistic pathogen produces and
substantial activation of the genetically engineered QS secretes two lipases LipA and LipC,[10] in addition to an
monitor MH205, compared with medium devoid of added outer membrane-located esterase EstA.[11] We assessed the
lipase and unsupplemented media (Figure 1). viability (colony forming units mL 1) of P. aeruginosa bacteria

440 www.angewandte.org  2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2014, 53, 439 –441
 Angewandte
Chemie

mode of delivery, dosage of antibiotics at the site of infection,

and development of antibiotic resistance. The self-regulating
system provides the basis for the development of device-
relevant polymeric materials, which only release antibiotics in
dependency of the titer of bacteria surrounding the medical
device. In the future, we hope that this concept can be
developed to construct anti-biofim or repellent coatings for
indwelling devices which can more effectively clear infections
caused by pathogenic bacteria.

Received: September 10, 2013

Revised: October 18, 2013
Published online: November 29, 2013

.Keywords: lipases · polymers · quorum sensing ·

solid-phase synthesis · surface chemistry

Figure 2. Viability of P. aeruginosa wild-type strain and lipA lipC estA

[1] a) K. K. Lai, S. A. Fontecchio, Am. J. Infect. Control 2002, 30,
triple mutant in presence of ciprofloxacin beads 8 (CFU = colony-
221 – 225; b) M. Lee, J. F. Donovan, J. Endourol. 2002, 16, 241 –
forming unit).
244; c) R. M. Donlan, J. W. Costerton, Clin. Microbiol. Rev.
2002, 15, 167 – 193.
[2] a) N. Barraud, B. G. Kardak, N. R. Yepuri, R. P. Howlin, J. S.
in cultures supplemented with beads 8. As a negative control Webb, S. N. Faust, S. Kjelleberg, S. A. Rice, M. J. Kelso, Angew.
Chem. 2012, 124, 9191 – 9194; Angew. Chem. Int. Ed. 2012, 51,
we used a P. aeruginosa lipA lipC estA triple mutant, which is
9057 – 9060; b) P. S. Stewart, J. W. Costerton, Lancet 2001, 358,
unable to produce the extracellular LipA, LipC, and EstA 135 – 138.
lipolytic enzymes. The killing kinetics of the P. aeruginosa [3] a) K. Lewis, Antimicrob. Agents Chemother. 2001, 45, 999 – 1007;
wild-type strain in the presence of pure ciprofloxacin b) M. R. Parsek, P. K. Singh, Annu. Rev. Microbiol. 2003, 57,
(0.35 mg mL 1) suggested that the concentration of ciproflox- 677 – 701.
acin released from the beads approaches this value. In [4] a) J.-Y. Wach, S. Bonazzi, K. Gademann, Angew. Chem. 2008,
comparison, the mean concentration of ciprofloxacin in 120, 7232 – 7235; Angew. Chem. Int. Ed. 2008, 47, 7123 – 7126;
b) F. Baneyx, D. T. Schwartz, Curr. Opin. Biotechnol. 2007, 18,
serum 12 h after dosing with 500 mg is typically around
312 – 317; c) M. Yoshinari, T. Kato, K. Matsuzaka, T. Hayakawa,
0.2 mg mL 1. As shown in Figure 2, the wild-type strain was K. Shiba, Biofouling 2010, 26, 103 – 110; d) P.-H. Chua, K.-G.
completely killed in the presence of beads 8 within 4 h, Neoh, E.-T. Kang, W. Wang, Biomaterials 2008, 29, 1412 – 1421;
whereas the population of the lipase-defective mutant was e) Z. Shi, K. G. Neoh, E. T. Kang, C. Poh, W. Wang, J. Biomed.
only insignificantly decreased in the presence of beads 8. Mater. Res. 2008, 86, 865 – 872; f) L. C. Gerber, F. M. Koehler,
The polymer material modified with amide-bonded R. N. Grass, W. J. Stark, Angew. Chem. 2012, 124, 11455 – 11458;
Angew. Chem. Int. Ed. 2012, 51, 11293 – 11296.
ciprofloxacin, so that the drug could not be released by the
[5] M. L. Vasil, L. M. Graham, R. M. Ostroff, V. D. Shortridge, A. I.
lipase or growth medium, did not display any antibiotic effect, Vasil, Antibiot. Chemother. 1991, 44, 34 – 47.
neither against the wild-type strain, nor for the lipase- [6] a) J. S. Dickschat, Nat. Prod. Rep. 2009, 27, 343 – 369; b) H.-S.
defective mutant (see the Supporting Information). Joo, M. Otto, Chem. Biol. 2012, 19, 1503 – 1513; b) C. Fuqua,
In the present study we have shown the feasibility of E. P. Greenberg, Nat. Rev. Mol. Cell Biol. 2002, 3, 685 – 695.
bacteria-triggered release of bioactive compounds, including [7] a) T. Bjarnsholt, M. Givskov, Philos. Trans. R. Soc. London Ser.
QS signals and antimicrobial drugs, from chemically modified B 2007, 362, 1213 – 1222; b) T. Praneenararat, A. G. Palmer,
H. E. Blackwell, Org. Biomol. Chem. 2012, 10, 8189 – 8199;
surfaces. The active compound precursor (or pro-drug) can be
c) W. R. J. D. Galloway, J. T. Hodgkinson, S. Bowden, M. Welch,
synthesized directly on the surface via standard solid-phase D. R. Spring, Trends Microbiol. 2012, 20, 449 – 458; d) C. A.
synthesis techniques. A key structural feature of the precursor Lowery, N. T. Salzameda, D. Sawada, G. F. Kaufmann, K. D.
is the presence of a cleavable anhydride bond which can be Janda, J. Med. Chem. 2010, 53, 7467 – 7489.
hydrolyzed in bacterial media by extracellular lipases. The [8] For complete synthetic protocols see Supporting Information.
lipase is eventually provided by the amassing bacteria as they [9] Enzymatic hydrolysis of substrates bound to the PEG-derived
approach the surface at the stage of early biofilm formation. materials has been widely utilized in one bead-two compound
assays. For an overview, see: M. Meldal, Curr. Opin. Chem. Biol.
Many compounds with a diverse range of biological activities
2004, 8, 238 – 244.
may in principle be released, including compounds that will [10] F. Rosenau, K. E. Jaeger, Biochimie 2000, 82, 1023 – 1032; A.
either stimulate or block QS, and subsequently sensitize Martinez, P. Ostrovsky, D. N. Nunn, Mol. Microbiol. 1999, 34,
bacteria to the action of the immune system and antibiotics. 317 – 326.
The approach may address some of the key challenges for the [11] S. Wilhelm, J. Tommassen, K. E. Jaeger, J. Bacteriol. 1999, 181,
treatment of bacterial infectious disease, including both the 6977 – 6986.

Angew. Chem. Int. Ed. 2014, 53, 439 –441  2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.angewandte.org 441