Soumya Palliyil, Christina Downham, Ian Broadbent, Keith Charlton and Andrew J. Porter Appl. Environ. Microbiol. 2014, 80(2):462. DOI: 10.1128/AEM.02912-13. Published Ahead of Print 1 November 2013.

Updated information and services can be found at: http://aem.asm.org/content/80/2/462 These include: SUPPLEMENTAL MATERIAL REFERENCES

CONTENT ALERTS

Supplemental material This article cites 34 articles, 17 of which can be accessed free at: http://aem.asm.org/content/80/2/462#ref-list-1 Receive: RSS Feeds, eTOCs, free email alerts (when new articles cite this article), more»

Information about commercial reprint orders: http://journals.asm.org/site/misc/reprints.xhtml To subscribe to to another ASM Journal go to: http://journals.asm.org/site/subscriptions/

Downloaded from http://aem.asm.org/ on January 10, 2014 by UNIVERSITAETS- UND LANDESBIBLIOTHEK DUESSELDORF

High-Sensitivity Monoclonal Antibodies Specific for Homoserine Lactones Protect Mice from Lethal Pseudomonas aeruginosa Infections

Soumya Palliyil,a Christina Downham,a Ian Broadbent,b Keith Charlton,c Andrew J. Portera Scottish Biologics Facility, University of Aberdeen, Foresterhill, Aberdeen, United Kingdoma; Department of Management, Aberdeen Business School, Robert Gordon University, Aberdeen, United Kingdomb; Scotia Biologics Ltd., Foresterhill, Aberdeen, United Kingdomc

A number of bacteria, including pathogens like Pseudomonas aeruginosa, utilize homoserine lactones (HSLs) as quorum sensing (QS) signaling compounds and engage in cell-to-cell communication to coordinate their behavior. Blocking this bacterial communication may be an attractive strategy for infection control as QS takes a central role in P. aeruginosa biology. In this study, immunomodulation of HSL molecules by monoclonal antibodies (MAbs) was used as a novel approach to prevent P. aeruginosa infections and as tools to detect HSLs in bodily fluids as a possible first clue to an undiagnosed Gram-negative infection. Using sheep immunization and recombinant antibody technology, a panel of sheep-mouse chimeric MAbs were generated which recognized HSL compounds with high sensitivity (nanomolar range) and cross-reactivity. These MAbs retained their nanomolar sensitivity in complex matrices and were able to recognize HSLs in P. aeruginosa cultures grown in the presence of urine. In a nematode slow-killing assay, HSL MAbs significantly increased the survival of worms fed on the antibiotic-resistant strain PA058. The therapeutic benefit of these MAbs was further studied using a mouse model of Pseudomonas infection in which groups of mice treated with HSL-2 and HSL-4 MAbs survived, 7 days after pathogen challenge, in significantly greater numbers (83 and 67%, respectively) compared with the control groups. This body of work has provided early proof-of-concept data to demonstrate the potential of HSL-specific, monoclonal antibodies as theranostic clinical leads suitable for the diagnosis, prevention, and treatment of life-threatening bacterial infections.

P

seudomonas aeruginosa is an opportunistic bacterial pathogen which causes life-threatening infections in immunocompromised individuals, including cystic fibrosis (CF), mechanically ventilated or catheterized, neutropenic, and burn patients. P. aeruginosa is the second most common cause of health care-associated pneumonia and the leading cause of pneumonia in pediatric patients in intensive care units (1, 2). It also dominates the Gram-negative group of bacteria which cause urosepsis, the most severe clinical manifestation of urinary tract infection (3). P. aeruginosa infections are currently treated with fluoroquinolones, aminoglycosides, carbapenems, and cephalosporins. However, this bacterium is notorious among clinicians for its ability to develop resistance against conventional antibiotics (4). Extensive studies have revealed two cell-cell communication, or quorum sensing (QS), pathways operating in P. aeruginosa— the las and rhl systems. Together they control the expression of extracellular virulence factors associated with Pseudomonas infection and also biofilm formation. In P. aeruginosa, QS is controlled by two low-molecular-weight autoinducer compounds called homoserine lactones (HSLs): N-(3-oxododecanoyl)-homoserine lactone (3-oxo-C12-HSL) and N-(butyryl)-homoserine lactone (C4-HSL) (5). Extracellular virulence factors such as flagellum, adhesion factors, including alginate, exotoxin A, exoenzyme S, elastases (LasA and LasB), alkaline protease, rhamnolipids, hydrogen cyanide, and phospholipase C are known to be controlled by the las and rhl systems, and the bacteria produce these factors in a cell density-dependent manner (6, 7). Apart from controlling the expression of virulence factors, 3-oxo-C12-HSL itself exerts significant immunomodulatory effects on mammalian immune responses (8). The extracellular distribution of QS compounds would appear to make them ideal targets for anti-infective therapy, since the evolutionary pressure on bacteria to develop resistance will be

462

aem.asm.org

limited. Several groups have used natural and synthetic HSL analogues or enzymes such as acylases and lactonases to interfere with las and rhl signaling cascades or degrade AHL signaling compounds (9–11). Monoclonal antibody (MAb)-mediated control of bacterial quorum sensing may offer exquisite target specificity and less off-target cytotoxicity (4). Quorum quenching using antibodies would not only block the signaling cascade and virulence factor production but also neutralize 3-oxo-C12-HSL, a major immunomodulatory component of P. aeruginosa infections (12). Although antibodies against 3-oxo-C12-HSL have demonstrated modest efficacy in previous studies, these antibodies have low sensitivities (often micromolar) to what are very challenging antigens (small size and simple chemical structure). Furthermore, existing monoclonal antibodies show little or no cross-reactivity with related HSL compounds (13). Antibodies that cross-react with C4-HSL would provide an added therapeutic advantage as they may also prevent the establishment of rhl QS system effects in the absence of the las system. In this study, an immunization approach was adopted to develop antibodies with increased affinity and improved sensitivity (100 to 1,000 times) compared to monoclonal antibodies isolated previously and even in complex matrices such as urine (13; K. Charlton and A. J. Porter, U.S. patent application 20130045208).

Applied and Environmental Microbiology

Received 29 August 2013 Accepted 29 October 2013 Published ahead of print 1 November 2013 Address correspondence to Soumya Palliyil, [email protected]. Supplemental material for this article may be found at http://dx.doi.org/10.1128 /AEM.02912-13. Copyright © 2014, American Society for Microbiology. All Rights Reserved. doi:10.1128/AEM.02912-13

p. 462– 469

January 2014 Volume 80 Number 2

Downloaded from http://aem.asm.org/ on January 10, 2014 by UNIVERSITAETS- UND LANDESBIBLIOTHEK DUESSELDORF

High-Sensitivity Monoclonal Antibodies Specific for Homoserine Lactones Protect Mice from Lethal Pseudomonas aeruginosa Infections

Monoclonal Antibodies for Pseudomonas Infection

A

O

Pan 1

3-oxo-C12 BSA (50 μg/ml)

O

TEA Eluon (100 mM)

O

B

O

Pan 2

NH CH3

O O

C

O

N-acyl-C12 BSA (5 μg/ml)

FA Eluon – 1 μM 3-oxo-C12-HSL

TEA Eluon (100 mM)

OH

NH

O

CH3

Pan 3

3-OH-C12 BSA (100 ng/ml)

FA Eluon – 10 μM C4-HSL

N-acyl-C12 BSA (100 ng/ml)

FA Eluon – 10 μM C4-HSL

FIG 1 Chemical structures of three HSL compounds used for sheep immunization. (A) N-Acyl-C12-HSL; (B) 3-OH-C12-HSL; (C) 3-oxo-C12-HSL. In P. aeruginosa the autoinducers 3-oxo-C12-HSL and C4-HSL share similar lactone ring structures but have different subgroups at the third carbon position and also vary in the length of their acyl side chains. The three HSL antigens synthesized for this work had the same lactone ring structure and various functional groups at the C-3 position in order to drive the immune response toward cross-reactive antibodies. These antigens were conjugated to the carrier protein TG and BSA for sheep immunization and library panning.

The protective effect of anti-quorum sensing antibodies was investigated in a slow-killing model of the nematode worm Caenorhabditis elegans. Finally, their ability to prolong survival and reduce viable bacterial load in the lungs was studied using a mouse model of P. aeruginosa infection. MATERIALS AND METHODS Bacterial strains. P. aeruginosa strains PAO1 and PA14 have been described in references 15 and 16, respectively. For the C. elegans slow-killing assay and mouse model of survival studies, P. aeruginosa clinical isolate strain 04.232058R (kindly donated by Tim Mitchell, University of Glasgow; referred as PA058 hereafter) was used. Antibiotic susceptibilities of this strain were studied previously in our laboratory, and it was found to be resistant to ciprofloxacin and tazocin and sensitive to gentamicin, ceftazidime, and meropenem. Primers. Ovine antibody constant-region primers used for firststrand cDNA synthesis, ovine heavy-chain variable-region (VH) 5= primers, ovine lambda chain variable-region (V␭) 3= primers, ovine kappa chain variable-region (V␬) 3= primers, and cloning primers with SfiI and NotI restriction sites were published in reference 17. All primers used for recombinant antibody library construction and sheep-mouse chimeric IgG conversion are given in the supplemental material. HSL compounds and conjugates. Free N-acyl-C12-HSL (N-dodecanoyl DL-homoserine lactone; catalog no. 17247) and C4-HSL (Nbutyryl-DL homoserine lactone; catalog no. 09945) were purchased from Sigma-Aldrich. Free 3-oxo-C12-HSL and 3-OH-C12-HSL were custom synthesized by Salford Ultrafine Chemicals and Research Ltd., Manchester, United Kingdom (presently known as SAFC, a division of SigmaAldrich Corporation). A series of synthetic analogues of HSL compounds with a carboxylic group at the end of the acyl side chain were synthesized and coupled via the carboxylic group to bovine serum albumin (BSA) and bovine thyroglobulin (TG) in our laboratory using 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC) (18). Sheep immunization and antibody library construction. Two Welsh breed sheep were hyperimmunized with a mixture of three HSL-TG conjugates—N-acyl-C12-HSL-TG, 3-oxo-C12-HSL-TG, and 3-OH-C12HSL-TG (Fig. 1)— by Micropharm Ltd., Wales, United Kingdom. For primary immunization, 500 ␮g of HSL conjugate mixture was administered per sheep, and for subsequent boosts, 250 ␮g/sheep was used.

January 2014 Volume 80 Number 2

FIG 2 Schematic representation of the panning strategy used for isolating cross-reactive HSL clones from an immunized antibody phage display library. FA, free antigen; TEA, triethyl amine.

Peripheral blood lymphocytes (PBLs) were prepared from sheep blood using Accuspin system Histopaque 1077 columns (Sigma; catalog no. A7054) according to the manufacturer’s instructions. Total RNA extraction, cDNA synthesis, and amplification of antibody variable heavy- and light-chain genes were performed as per published methods (17). For linking of antibody genes, VH and V␭/V␬ PCR products were joined by restriction enzyme digestion and ligation using the restriction sites AscI for the heavy chain and MluI for the light chain at the linker region incorporated through PCR design. The cloning sites SfiI and NotI were incorporated to the ligated DNA as reported previously (17), and the phagemid vector pHEN 2 (19) was used for single-chain variable fragment (scFv) library construction. The resultant VH-V␭ and VH-V␬ libraries were rescued separately through helper phage infection following the method described in reference 20. Selection and screening of HSL binders using phage display. Biopanning, monoclonal phage binding enzyme-linked immunosorbent assay (ELISA), and phage competition ELISA were performed according to the methods published in reference 20. The panning strategy was designed to drive selection toward the enrichment of cross-reactive clones as shown schematically in Fig. 2. The coating concentration of HSL conjugates— N-acyl-C12-BSA, 3-oxo-C12-BSA, and 3-OH-C12-BSA—and carrier protein BSA for ELISA was 1 ␮g/ml. The clones which showed reduced binding toward HSL conjugates in the presence of free HSL compounds (phage competition ELISA) were sequenced and reformatted into sheep-mouse chimeric MAbs. Conversion of HSL scAbs into sheep-mouse chimeric MAbs. VH and V␭ genes from six HSL-positive clones (numbered HSL-1 to HSL-6) were amplified using variable-region-specific primers (see the supplemental material). Purified V␭ regions were joined with respective murine lambda 1 constant-region (MuC␭1) by overlapping PCR. Joined PCR products were inserted into one vector mammalian expression system, pHEE, using BamHI-PstI cloning sites to generate HSL pHEE light-chain vectors (HSL V␭–MuC␭1). Similarly variable heavy chains of HSL-positive clones were amplified using V-region-specific forward and reverse primers and joined to the murine immunoglobulin gamma 1 (MuIgG1) constant region using MfeI and BsmBI cloning sites. The heavy chains of HSL clones which were successfully linked to MuIgG1 constant region were then cloned into respective pHEE–light chain vectors (HSL V␭-MuC␭1) using EcoRI-XbaI cloning sites and transformed electrocompetent XL1-Blue cells (Agilent Technologies). Expression of HSL MAbs in mammalian expression system. Dihydrofolate reductase (DHFR)-deficient CHO DG44 cells (Invitrogen catalog no. A11000-01) were maintained in Iscove’s modified Dulbecco’s me-

aem.asm.org 463

Downloaded from http://aem.asm.org/ on January 10, 2014 by UNIVERSITAETS- UND LANDESBIBLIOTHEK DUESSELDORF

CH3

Palliyil et al.

464

aem.asm.org

to 25°C. The numbers of dead worms on the plate were determined at 12-h time points and the count was repeated up to 96 h. P. aeruginosa PA058 survival study. Certified pathogen-free CD1 mice supplied by Charles River UK were infected with ⬃1.52 ⫻ 107 CFU of P. aeruginosa PA058 by intranasal (i.n.) administration under parenteral anesthesia (ketamine [100 mg/kg of body weight]-xylazine [6 mg/ kg]). Mice were treated with test MAbs—HSL-2 and HSL-4 —at 10 mg/kg or a vehicle-only control that was coadministered with the inocula i.n. A second treatment dose was delivered by intravenous (i.v.) injection 4 h postinfection. The comparator drugs gentamicin and ciprofloxacin were delivered at 10 mg/kg i.v. once and 150 mg/kg per os (p.o.) twice, respectively. Five mice per treatment group were euthanized for quantitative culture of lung tissue for organism burden 24 h postinfection; six mice per group were observed for survival up to 7 days postinfection. Statistics. Quantitative tissue burden data were analyzed using the Kruskal-Wallis statistical test and compared to the vehicle control group using StatsDirect. Survival estimates were determined by the KaplanMeier test.

RESULTS

Characterization of sheep polyclonal sera using elastase assay. Sheep immunization with HSL conjugates resulted in a high titer of anti-HSL polyclonal antibodies in immune sera that were capable of recognizing native HSL compounds with high affinity and sensitivity in competition ELISA (Table 1). The potential neutralizing activity of polyclonal sheep antibodies was analyzed in vitro using a QS-regulated elastolytic activity assay. When P. aeruginosa was grown in the presence of anti-HSL polyclonal antibodies, a marked reduction in elastase production was noticed (Fig. 3). The positive control, Hap5 MAb (Charlton and Porter, U.S. patent application 20130045208), also reduced the level of elastase production (Fig. 3). Biopanning of HSL-immunized library. Using PBLs from immunized sheep as starting material, VH-V␭ and VH-V␬ antiHSL phage display libraries were constructed in an scFv format. The cellulase linker used for this work was a modified version previously used successfully in our laboratory (26). Sequence analysis revealed 100% diversity in both lambda and kappa libraries, and the sizes were estimated at 2.8 ⫻ 107 and 1.3 ⫻ 107, respectively. The libraries were rescued using M13K07 helper phage separately, and following biopanning and monoclonal phage ELISA, 20 six clones were selected which showed reduced HSL conjugate binding in the presence of free HSL compounds. Six unique free antigen binders were identified and named HSL-1 to HSL-6. These six clones were converted into soluble single-chain antibody (scAb) format by cloning into the expression vector pIMS147 (27). The sensitivity and cross-reactivity of positive clones were determined by competition ELISA using subsaturating concentrations of HSL scAbs. All HSL scAbs showed high sensitivity for 3-oxo-C12-HSL, 3-OH-C12-HSL, and N-acyl-C12-HSL compounds (expressed as their IC50s in Table 1). HSL-1, HSL-4, and HSL-6 had subnanomolar sensitivities, with IC50s of 500 pM, 500 pM, and 400 pM, respectively, for 3-oxo-C12-HSL, which is the central QS compound in P. aeruginosa. All six HSL antibodies showed reduced binding to C4-HSL, which is the second most important autoinducer compound in the P. aeruginosa QS circuit. Generation of HSL MAbs in mammalian expression system and characterization by ELISA. HSL-1 to HSL-6 clones were converted into sheep-mouse IgG1-␭1 chimeric monoclonal antibodies and expressed transiently in COS-7 cells; HSL-2 and HSL-4

Applied and Environmental Microbiology

Downloaded from http://aem.asm.org/ on January 10, 2014 by UNIVERSITAETS- UND LANDESBIBLIOTHEK DUESSELDORF

dium (IMDM) with standard supplements and transfected with 20 ␮g of a plasmid DNA-containing antibody insert using Lipofectamine 2000 (Invitrogen; catalog no. 11668019) by following the manufacturer’s instructions. The transfectomas stably expressing recombinant HSL antibodies were selected following the method described in reference 21. The best antibody producers for HSL-2 and HSL-4 were then used to seed two-liter culture chambers (CELL STACK; Corning; catalog no. TKT-534-030L), and after 14 days, the culture supernatant was harvested and purified using Prosep-vA Ultra column chromatography (Millipore; catalog no. 115115824). ELISA-based characterization of positive MAbs. (i) Capture ELISA for MAb quantification. Capture ELISA for MAb quantification was performed using a modified method previously described (22). The capture antibody used was a 1-in-1,000 dilution goat anti-mouse IgG1 antibody (ABD Serotech; 107001), and the secondary antibody was a 1-in-1,000 dilution goat anti-mouse IgG1 horseradish peroxidase (HRP)-conjugated polyclonal antibody (ABD Serotech; 132P). Monoclonal mouse IgG1 (ABD Serotech, 5220-2959) was used as the standard antibody with a starting concentration of 1 ␮g/ml. (ii) MAb antigen binding and competition ELISA. MAb antigen binding and competition ELISA were performed using modified methods from reference 23. For MAb competition ELISA, double dilutions of free HSL solutions (N-acyl-C12, 3-oxo-C12, 3-OH-C12, and C4-HSL) were mixed with equal volumes of a subsaturating concentration of HSL MAbs (determined by binding ELISA) and preincubated at room temperature for 1 h. For 100% binding control, 1⫻ phosphate-buffered saline (PBS) was mixed with HSL MAbs. Secondary antibody binding and development were performed as described above. (iii) Detection of HSL compounds in urine culture filtrates. P. aeruginosa PA14 and PAO1 and Escherichia coli OP50 cultures were grown separately in 5 ml of fresh human urine for 3 days at 37°C with shaking. The cultures were spun down at 2,000 ⫻ g for 10 min and the resultant supernatants passed through 0.45-␮m filters. Double dilutions of culture supernatants in urine were mixed with equal volumes of a subsaturating concentration of HSL-2 MAb and used in a competition ELISA format. Elastase assay. Elastase production was measured using the published method (24), with some modifications. Briefly, a starter culture of P. aeruginosa PA14 at an optical density at 600 nm (OD600) of 0.05 was prepared from an overnight culture in Luria-Bertani (LB) medium. Duplicate 2-ml starter culture samples were added to separate tubes containing purified sheep polyclonal sera, preimmune sera, and control antibody Hap5 MAb at a final IgG concentration of 600 nM. For the assay negative control, an overnight-grown E. coli OP50 culture was used without addition of any polyclonal antibodies (non-elastase-producing strain). The cultures were grown at 37°C with shaking at 250 rpm until an OD600 of 1.8 to 2.0 was reached (early stationary phase), and 1 ml of the culture was centrifuged at 17,000 ⫻ g for 10 min for each sample. One milliliter of culture supernatant was mixed with 2 ml of elastin-Congo red suspension and incubated at 37°C for 16 to 18 h with shaking. The overnight samples were centrifuged as before, and absorbance of the supernatants at 495 nm was read against a blank (E. coli OP50). C. elegans slow-killing assay. The assay was performed using the wild-type Bristol N2 strain and following the published protocol (25). Separate Pseudomonas cultures were grown in LB medium containing 500 nM HSL-2 and HSL-4 MAbs until log phase. A small sample (10 ␮l) from each culture was spotted onto 3.5-cm agar plates containing nematode growth medium supplemented with extra peptone and previously treated with HSL MAbs. The plated cultures were allowed to grow for 24 h at 37°C. The following day, plates were treated with same dosage of MAbs and incubated at room temperature. Additional antibody was added periodically (12 h) throughout the incubation period. Controls were set up using E. coli OP50 and PA058 without any antibodies. A population of 40 to 50 young adult worms was added onto the center of each of the PA058 with or without antibody or E. coli OP50 plates in a drop of M9 and encouraged to feed on the bacterial lawn over a period of 3 to 4 days at 23

Monoclonal Antibodies for Pseudomonas Infection

IC50 for: Sera or antibody

3-Oxo-C12-HSL

3-OH-C12-HSL

N-Acyl-C12-HSL

C4-HSL

Sheep polyclonal sera HSL-1 scAb HSL-2 scAb HSL-3 scAb HSL-4 scAb HSL-5 scAb HSL-6 scAb HSL-2 MAb HSL-4 MAb Human MAb Hap5 (Charlton and Porter, U.S. patent application 20130045208) Mouse MAb RS2-1G9 (13)

20 nM 0.5 nM 1.5 nM 2.5 nM 0.5 nM 2.5 nM 0.4 nM 4 nM 1.5 nM 10 ␮M

180 nM 170 nM 170 nM 700 nM 200 nM 700 nM 40 nM 40 nM 40 nM CND

50 nM 700 nM 1.5 ␮M 5 ␮M 1 ␮M 3.5 ␮M 2 ␮M 400 nM 280 nM 7 ␮M

12 ␮M 100 ␮M ND ND 100 ␮M ND 100 ␮M 175 ␮M 130 ␮M ⬎750 ␮M

150 nM

CND

CND

⬎100 ␮M

a

Values for polyclonal sera and HSL antibodies used in this study are significantly lower (100- to 1,000-fold increased sensitivity) than those of similar monoclonal antibodies developed against the same targets by other groups: human MAb Hap5 (Charlton and Porter, U.S. patent application 20130045208) and mouse MAb RS2-1G9 (13). ND, not determined; CND, could not be determined.

Absorbance OD 450NM

MAbs were identified as high expressors as determined by capture ELISA (20 ng/ml and 2 ng/ml in cell culture supernatant, respectively) and therefore selected for stable cell line generation and further characterization as diagnostic and/or therapeutic leads. For HSL-2 and HSL-4 MAbs, the best clonal producers expressed antibodies in the region of 20 ␮g/ml. The concentrations of free 3-oxo-C12-HSL required to reduce HSL MAb binding by 50% were 4 nM (HSL-2) and 1.5 nM (HSL-4) (Fig. 4). These IC50s were slightly higher than their equivalent scAbs, although no considerable reduction in antigen sensitivities was observed following IgG conversion. Interestingly, the MAbs recognized free N-acyl-C12-HSL, 3-OH-C12-HSL, and C4HSL compounds with increased sensitivity compared to their scAb counterparts, indicating an improved cross-reactivity (Fig. 4 and 5). A comparison of the IC50s of HSL Abs in polyclonal, scAb, and MAb formats is given in Table 1. In order to explore the diagnostic potential of our sheepmouse chimeric MAbs, the ability of HSL-2 MAb to detect native autoinducer compounds in P. aeruginosa cultures grown in urine was assessed in a competition ELISA format. HSL-2 MAb successfully recognized autoinducer compounds in P. aeruginosa strains

Average Elastolyc acvity

2.5 2.0 1.5 1.0 0.5 0.0 PBS

Pre-Immune serum

Sheep 1 immune

Sheep 2 immune

Hap 5 mAb

E.coli OP50

FIG 3 Elastolytic activity determined by measuring the levels of soluble elastin-Congo red in P. aeruginosa PA14 in the presence of sheep polyclonal sera and a control antibody, Hap5 (K. Charlton and A. J. Porter, U.S. patent application 20130045208) (600 nM). Control levels of elastase were produced when cells were grown in the presence of PBS or preimmune serum. Data plotted represent the means from three assay repeats plus SEs.

January 2014 Volume 80 Number 2

PA14 and PAO1 (Fig. 6). E. coli strain OP50 was used as a negative control and purified free 3-oxo-C12-HSL diluted in urine as a positive control. In vivo efficacy studies. The nematode worm C. elegans was selected as an excellent initial in vivo model for studying the protective effect of HSL MAbs during P. aeruginosa infection. P. aeruginosa can cause lethal infections of C. elegans, and the slowkilling assay is dependent upon the action of HSL-mediated quorum sensing mechanisms (28). Over a period of 96 h, the survival rates of C. elegans treated with 500 nM HSL-2 and HSL-4 MAbs were increased from 15% to 53% and 60%, respectively (Fig. 7). P. aeruginosa strain PA058 is ciprofloxacin resistant, and therefore, no increase in survival rate was noticed in this group or in the presence of an unrelated MAb and PBS control. The in vivo efficacy of HSL-2 and HSL-4 MAbs was further assessed in a nonneutropenic lung model of mice infected with P. aeruginosa PA058 using survival endpoints after treatment delivered as monotherapy. A moderate level of localized P. aeruginosa infection was established via intranasal administration of PA058 suspension into the nostrils of mice under parenteral anesthesia, with a geometric mean burden in the vehicle group of ⬃1.7 ⫻ 104 CFU/g of lung tissue at 24 h postinfection. All vehicletreated animals succumbed to disease by 48 h postinfection. Mice treated with HSL-2 and HSL-4 MAbs as monotherapy had significantly prolonged survival times (83% and 67% survival 7 days postinfection, respectively; P ⫽ 0.0022 by StatsDirect MannWhitney U test comparison of survival) compared with those of vehicle-only-treated mice (0% survival 2 days postinfection) (Fig. 8). The comparator drug gentamicin delivered once at 10 mg/kg i.v. also significantly prolonged survival, with 100% survival 7 days postinfection (P ⫽ 0.0022 by StatsDirect Mann-Whitney U test comparison of survival) compared to that of vehicle-onlytreated mice. However, the comparator ciprofloxacin did not show statistically significant improvement in survival when delivered as monotherapy (40% survival 7 days postinfection). The poor efficacy of ciprofloxacin is in line with expectations based on the in vitro MIC (10 ␮g/ml), which is higher than the susceptibility breakpoint.

aem.asm.org 465

Downloaded from http://aem.asm.org/ on January 10, 2014 by UNIVERSITAETS- UND LANDESBIBLIOTHEK DUESSELDORF

TABLE 1 Comparison of IC50s for polyclonal sera, HSL scAbs (HSL-1 to -6), and HSL-2 and HSL-4 MAbs specific for different subgroups of native HSL compounds in a competition ELISA formata

Palliyil et al.

110 100

90

% Relative absorbance

% Relative absorbance

100 80 70 60 50 40 30 20 10

90 80 70 60 50 40 30 20 10 0

0 0.1

1

10

100

0.1

1000

1

10

100

1000

FIG 4 Competitive inhibition of HSL-2 (䡲) and HSL-4 (Œ) MAb binding to 3-oxo-C12-BSA conjugate in the presence of free 3-oxo-C12-HSL (A) and 3-OH-C12-BSA conjugate in the presence of free 3-OH-C12-HSL (B), in a range of concentrations starting from 200 nM. Data points represent means of duplicate values obtained for each sample.

DISCUSSION

110 100 90 80 70 60 50 40 30 20 10 0 0.01

of this approach is that the lactam analogue is a nonnatural antigen, and in most cases antibodies resulting from immunization fail to recognize native lactone compounds. Our approach, however, was to use the native lactone ring, but with different subgroups at the third carbon position (N-acyl-C12-, 3-oxo-C12-, and 3-OH-C12-HSL). Two sheep were hyperimmunized with a conjugate mixture representing the three HSL subgroups to enhance the chance of eliciting an antibody response. The acyl side chain present in all three compounds (Fig. 1) was used as a natural spacer for conjugation, and therefore, generation of any interface binders which recognized the coupling position of the hapten carrier protein was avoided. In addition, by conjugating carrier proteins to the acyl side chain, the lactone ring structure can be presented in a correct orientation, so that the immune response is mostly directed toward this common epitope of interest. Sheep were chosen as the host animal for immunization, as they have been shown to generate high-affinity antibodies against haptenic targets (20) and sheep polyclonal antibodies have been used as specific high-affinity immunologic probes for analytical and clinical purposes for many years (35). The sensitivities of sheep polyclonal antibodies to native HSL compounds were far superior to the published mouse monoclonal antibody sensitivity toward the same targets (13), and the neutralizing effect of these polyclonal antibodies was first demonstrated in an elastase assay (Fig. 3).

110 100

% Relative absorbance

% Relative absorbance

In P. aeruginosa chronic infections, autoinducer concentrations in patient samples vary among different strains, and accurate values are still unknown. However, in one report, the levels of extractable 3-oxo-C12-HSL and C4-HSL in the sputum of CF patients were in the nanomolar range: up to 22 nM for 3-oxo-C12-HSL and 0 to 5 nM for C4-HSL (29). This is considerably lower than the concentrations required for exerting cytotoxicity in mammalian cells (10 to 50 ␮M) (30) and also the levels found in planktonic P. aeruginosa cultures (1 ␮M) (31) and in vitro biofilms (600 ␮M) (32). It has been widely accepted that antibodies with subnanomolar affinity exhibit improved therapeutic efficacy, especially if the targets to be bound are present in low concentrations (33). Clearly, nanomolar levels of HSL compounds in patient lungs cannot be ruled out, although these lower levels observed might have been influenced analytically by factors such as the unstable nature of these compounds in the extracellular environment, restricted movement of the 3-oxo-C12-HSL out of the biofilm, heterogeneity in biofilm distribution, and differential lasI and/or rhlI gene expression (29). In a separate study carried out with mice, Debler and coworkers replaced the chemically labile lactone ring by a more stable lactam moiety for immunization and production of mouse monoclonal antibodies against AHL compounds (34). The disadvantage

0.1

1

10

100

90 80 70 60 50 40 30 20 10 0 1

10

100

1000

10000

FIG 5 Competitive inhibition of HSL-2 (䡲) and HSL-4 (Œ) MAb binding to N-acyl-C12-BSA conjugate in the presence of free N-acyl-C12-HSL (A) and free C4-HSL (B), in a range of concentrations starting from 50 ␮M and 10 mM, respectively. Data points represent means of duplicate values obtained for each sample.

466

aem.asm.org

Applied and Environmental Microbiology

Downloaded from http://aem.asm.org/ on January 10, 2014 by UNIVERSITAETS- UND LANDESBIBLIOTHEK DUESSELDORF

110

Monoclonal Antibodies for Pseudomonas Infection

120

100

10

1

100

Percentage survival

3oxo C12 HSL PAO1 PA14 E. coli

10

100

80 60 40 20

1000

0

FIG 6 Detection of HSL compounds using HSL-2 MAb in P. aeruginosa culture filtrates grown in urine. Competitive inhibition of HSL-2 MAb binding to 3-oxo-C12-BSA conjugate was observed for PAO1 and PA14 culture filtrates. Purified free 3-oxo-C12-HSL was used as a positive control and E. coli OP50 as a negative control. Data points represent means of duplicate values obtained for each sample.

Our MAb selection strategy was designed to drive enrichment toward, sensitive, cross-reactive phage antibody clones (26). The IC50s obtained from indirect competition ELISA using native 3-oxo-C12-HSL, 3-OH-C12-HSL, N-acyl-C12-HSL, and C4-HSL compounds suggested that the sensitivities of HSL MAbs were not reduced after IgG conversion. These values are quite impressive, considering the chemical nature of these lipid-like compounds, which possess a small head-like structure and lack critical antigenic features such as aromaticity or charge. The IC20 (concentra-

120

Percentage survival

100 80 60 40 20 0 0

20

40

60

80

100

Hours post infection E.coli OP50 HSL-2 mAb HSL-4 mAb Non-specific mAb Vehicle Ciprofloxacin

FIG 7 Kaplan-Meier plot showing the protective effect of HSL MAbs in a nematode model of P. aeruginosa slow-killing assay. HSL MAbs HSL-2 and HSL-4 at 500 nM increased the survival rate of C. elegans feeding on P. aeruginosa PA058 by 53% and 60%, respectively, compared with the control groups. Nematode groups which received the same concentration of nonspecific MAb, vehicle (phosphate-buffered saline; pH 7.4), and ciprofloxacin showed survival rates of 18%, 30%, and 18%, respectively. All nematode worms fed on E. coli OP50 survived during the entire course of the experiment.

January 2014 Volume 80 Number 2

0

1

2

3

4

5

6

7

Days post infection Vehicle Gentamicin HSL-2 mAb HSL-4 mAb Ciprofloxacin

FIG 8 Kaplan-Meier plot showing the survival of mice treated with 10 mg/kg of HSL-2 and HSL-4 MAbs in a lung infection model of P. aeruginosa PA058. HSL-2 and HSL-4 MAbs increased the survival by 83% and 67%, respectively, 7 days postinfection compared with the control groups. Mice treated with a vehicle (0.9% saline) and the comparator drug ciprofloxacin at 150 mg/kg showed survival rates of 0% and 40%, respectively. The comparator drug gentamicin delivered at 10 mg/kg also significantly prolonged survival, by 100%.

tion of free HSLs required to reduce the binding of antibodies to a competing antigen by 20%) could be considered the limit of detection for a particular antibody/antigen assay. Importantly, the IC20s of HSL-2 and HSL-4 antibodies (1.5 nM and 0.6 nM, respectively) for the important diagnostic class of homoserine lactones (3-oxo-C12-HSL) are much lower (100 to 1,000 times) than the extractable levels of HSL compounds found in many P. aeruginosa cultures (31). The IC50s revealed improved cross-reactivities of HSL MAbs (Table 1); however, the cross-reactivity toward the second autoinducer, C4-HSL, remained poor. The reported affinity of mouse MAb RS2-1G9 against 3-oxo-C12-HSL was 150 nM as determined by indirect competition ELISA (13). The authors noticed a 1,000fold-reduced sensitivity of this antibody for C4-HSL. Our HSL MAbs also showed good cross-reactivity toward 3-OH-C12-HSL and reduced sensitivity toward C4-HSL, which lacks a functional group at the third carbon position and has a reduced number of methylene units in its side chain. Typically for antibody neutralization, Fc-mediated host effector functions are an important component of in vivo efficacy. Here the mode of action is based around antibody-mediated scavenging of HSL compounds causing suppression of virulence factors and other inflammatory responses. To demonstrate this hypothesis we first used the nematode C. elegans infection model in which P. aeruginosa slow killing is mediated by the accumulation of bacterial load within the intestinal lumen of the nematode. Pseudomonas lasR mutants are less pathogenic in this model, as they fail to accumulate in the gut and produce reduced levels of elastase and protease (28). In our study the significant increase in survival rate in the presence of HSL-2 and HSL-4 MAbs is similar to the defective slow killing observed in lasR mutant strains.

aem.asm.org 467

Downloaded from http://aem.asm.org/ on January 10, 2014 by UNIVERSITAETS- UND LANDESBIBLIOTHEK DUESSELDORF

Relative absorbance %

Culture filtrate dilution 1/x

1000 110 100 90 80 70 60 50 40 30 20 10 0 1

Palliyil et al.

Parameter

IC20 (nM)

IC50 (nM)

HSL-2 MAb in PBS HSL-2 MAb in urine

1.5 5

4 10

a HSL-2 MAb retained functional recognition of its antigen in the presence of urine, with very little reduction in sensitivity observed.

In a nonneutropenic lung model of mice infected with P. aeruginosa PA058, HSL-2 and HSL-4 MAb monotherapy demonstrated significant efficacy, prolonging survival by 83% and 67%, respectively. However, this efficacy was not reflected in 24-h lung burdens, as there was no significant reduction in bacterial counts in mice infected with P. aeruginosa PA058 24 h postinfection (results not shown). This may reinforce the view that HSL-specific antibodies protect mice from P. aeruginosa infections possibly through antibody-mediated scavenging of HSL compounds and without necessarily affecting bacterial numbers. HSL scavenging may disrupt bacterial communication to such an extent that the bacteria are “held” as a benign or nonpathogenic phenotype. This model was performed in mice with fully functional immune systems, so following survival from the acute infectious challenge the disease may resolve through clearance of bacteria by multiple processes, including lung macrophages. Similar findings were reported by Miyairi and coworkers; in their study, active immunization using 3-oxo-C12-HSL conjugated to the carrier protein BSA in a P. aeruginosa lung infection model yielded high titers of immunogen-specific antibodies which conferred a modest survival benefit (30%) without reducing the pulmonary bacterial loads in immunized and nonimmunized groups (36). To conclude, using a combination of sheep immunization and phage antibody display library construction and selection, highsensitivity monoclonal antibodies were generated toward homoserine lactones, especially 3-oxo-C12-HSL. These antibodies have the potential to block quorum sensing signaling in P. aeruginosa and thereby regulate the expression of virulence factors involved in bacterial pathogenesis. Due to the lack of antibiotic cytotoxicity, HSL monoclonal antibodies have the drug class pedigree to provide a novel preventative for life-threatening P. aeruginosa infections in groups such as CF patients. The absence of inherent bactericidal activity may prove highly advantageous, as this therapeutic approach may not apply the same selection pressures on bacteria to develop resistance-enhancing cotherapy approaches. The monoclonal antibodies showed at least 100-fold improvement in sensitivity for P. aeruginosa autoinducer compounds in biochemical assays, compared to any other published antibodies raised to the same target (Table 1). An immunoassay-based diagnostic system exploiting the high sensitivity of HSL MAbs could be developed to detect the presence of specific markers of infection (homoserine lactones) in bodily fluids such as urine (Table 2). This improvement in MAb sensitivity is also believed to have had a corresponding positive impact on therapeutic efficacy in vivo, apparently blocking bacterial infection. Further research with lead compounds will better determine the potential of HSL monoclonal antibodies to prevent infections caused by antibiotic-resistant strains of P. aeruginosa using monotherapy or, more likely, cotherapy regimens.

468

aem.asm.org

ACKNOWLEDGMENTS This work was supported by the Knowledge Transfer Partnership UK, Pfizer Inc., Scottish Universities Life Sciences Alliance, and a University of Aberdeen Knowledge Transfer Grant. We thank Euprotec Ltd. Manchester, United Kingdom, for performing in vivo mouse efficacy studies.

REFERENCES 1. Richards MJ, Edwards JR, Culver DH, Gaynes RP. 1999. Nosocomial infections in pediatric intensive care units in the United States. National Nosocomial Infections Surveillance System. Pediatrics 103:e39. 2. Gaynes R, Edwards JR. 2005. Overview of nosocomial infections caused by gram-negative bacilli. Clin. Infect. Dis. 41:848 – 854. http://dx.doi.org /10.1086/432803. 3. Kalra OP, Raizada A. 2009. Approach to a patient with urosepsis. J. Glob. Infect. Dis. 1:57– 63. http://dx.doi.org/10.4103/0974-777X.52984. 4. Palliyil S, Broadbent ID. 2009. Novel immunotherapeutic approaches to the treatment of infections caused by Gram-negative bacteria. Curr. Opin. Pharmacol. 9:566 –570. http://dx.doi.org/10.1016/j.coph.2009.07.007. 5. Passador L, Cook JM, Gambello MJ, Rust L, Iglewski BH. 1993. Expression of Pseudomonas aeruginosa virulence genes requires cell-tocell communication. Science 260:1127–1130. http://dx.doi.org/10.1126 /science.8493556. 6. de Kievit TR, Iglewski BH. 2000. Bacterial quorum sensing in pathogenic relationships. Infect. Immun. 68:4839 – 4849. http://dx.doi.org/10.1128 /IAI.68.9.4839-4849.2000. 7. Whiteley M, Lee KM, Greenberg EP. 1999. Identification of genes controlled by quorum sensing in Pseudomonas aeruginosa. Proc. Natl. Acad. Sci. U. S. A. 96:13904 –13909. http://dx.doi.org/10.1073/pnas.96.24.13904. 8. Telford G, Wheeler D, Williams P, Tomkins PT, Appleby P, Sewell H, Stewart GS, Bycroft BW, Pritchard DI. 1998. The Pseudomonas aeruginosa quorum-sensing signal molecule N-(3-oxododecanoyl)- L homoserine lactone has immunomodulatory activity. Infect. Immun. 66: 36 – 42. 9. Givskov M, de Nys R, Manefield M, Gram L, Maximilien R, Eberl L, Molin S, Steinberg PD, Kjelleberg S. 1996. Eukaryotic interference with homoserine lactone-mediated prokaryotic signalling. J. Bacteriol. 178: 6618 – 6622. 10. Smith KM, Bu Y, Suga H. 2003. Induction and inhibition of Pseudomonas aeruginosa quorum sensing by synthetic autoinducer analogs. Chem. Biol. 10:81– 89. http://dx.doi.org/10.1016/S1074-5521(03)00002-4. 11. Lin YH, Xu JL, Hu J, Wang LH, Ong SL, Leadbetter JR, Zhang LH. 2003. Acyl-homoserine lactone acylase from Ralstonia strain XJ12B represents a novel and potent class of quorum-quenching enzymes. Mol. Microbiol. 47:849 – 860. http://dx.doi.org/10.1046/j.1365-2958.2003.03351.x. 12. Kaufmann GF, Park J, Mee JM, Ulevitch RJ, Janda KD. 2008. The quorum quenching antibody RS2-1G9 protects macrophages from the cytotoxic effects of the Pseudomonas aeruginosa quorum sensing signalling molecule N-3-oxo-dodecanoyl-homoserine lactone. Mol. Immunol. 45: 2710 –2714. http://dx.doi.org/10.1016/j.molimm.2008.01.010. 13. Kaufmann GF, Sartorio R, Lee SH, Mee JM, Altobell LJ, III, Kujawa DP, Jeffries E, Clapham B, Meijler MM, Janda KD. 2006. Antibody interference with N-acyl homoserine lactone-mediated bacterial quorum sensing. J. Am. Chem. Soc. 128:2802–2803. http://dx.doi.org/10.1021/ja0578698. 14. Reference deleted. 15. Klockgether J, Munder A, Neugebauer J, Davenport CF, Stanke F, Larbig KD, Heeb S, Schock U, Pohl TM, Wiehlmann L, Tummler B. 2010. Genome diversity of Pseudomonas aeruginosa PAO1 laboratory strains. J. Bacteriol. 192:1113–1121. http://dx.doi.org/10.1128/JB.01515 -09. 16. Rahme LG, Stevens EJ, Wolfort SF, Shao J, Tompkins RG, Ausubel FM. 1995. Common virulence factors for bacterial pathogenicity in plants and animals. Science 268:1899 –1902. http://dx.doi.org/10.1126/science .7604262. 17. Charlton KA, Moyle S, Porter AJ, Harris WJ. 2000. Analysis of the diversity of a sheep antibody repertoire as revealed from a bacteriophage display library. J. Immunol. 164:6221– 6229. 18. Hermanson G 1996 Bioconjugate techniques. Elsevier Science, San Diego, CA. 19. Hoogenboom HR, Griffiths AD, Johnson KS, Chiswell DJ, Hudson P, Winter G. 1991. Multi-subunit proteins on the surface of filamentous phage: methodologies for displaying antibody (Fab) heavy and light

Applied and Environmental Microbiology

Downloaded from http://aem.asm.org/ on January 10, 2014 by UNIVERSITAETS- UND LANDESBIBLIOTHEK DUESSELDORF

TABLE 2 Comparison of IC20s and IC50s for HSL-2 MAb specific for native 3-oxo-C12-HSL compound when the assay was performed in PBS and in urinea

Monoclonal Antibodies for Pseudomonas Infection

21.

22.

23.

24. 25.

26. 27.

28.

January 2014 Volume 80 Number 2

29. Erickson DL, Endersby R, Kirkham A, Stuber K, Vollman DD, Rabin HR, Mitchell I, Storey DG. 2002. Pseudomonas aeruginosa quorumsensing systems may control virulence factor expression in the lungs of patients with cystic fibrosis. Infect. Immun. 70:1783–1790. http://dx.doi .org/10.1128/IAI.70.4.1783-1790.2002. 30. Kravchenko VV, Kaufmann GF, Mathison JC, Scott DA, Katz AZ, Wood MR, Brogan AP, Lehmann M, Mee JM, Iwata K, Pan Q, Fearns C, Knaus UG, Meijler MM, Janda KD, Ulevitch RJ. 2006. N-(3-oxoacyl)homoserine lactones signal cell activation through a mechanism distinct from the canonical pathogen-associated molecular pattern recognition receptor pathways. J. Biol. Chem. 281:28822–28830. http://dx.doi.org /10.1074/jbc.M606613200. 31. Pearson JP, Passador L, Iglewski BH, Greenberg EP. 1995. A second N-acylhomoserine lactone signal produced by Pseudomonas aeruginosa. Proc. Natl. Acad. Sci. U. S. A. 92:1490 –1494. http://dx.doi.org/10.1073 /pnas.92.5.1490. 32. Charlton TS, de Nys R, Netting A, Kumar N, Hentzer M, Givskov M, Kjelleberg S. 2000. A novel and sensitive method for the quantification of N-3-oxoacyl homoserine lactones using gas chromatography-mass spectrometry: application to a model bacterial biofilm. Environ. Microbiol. 2:530 –541. http://dx.doi.org/10.1046/j.1462-2920.2000.00136.x. 33. Hoogenboom HR. 2005. Selecting and screening recombinant antibody libraries. Nat. Biotechnol. 23:1105–1116. http://dx.doi.org/10.1038/nbt1126. 34. Debler EW, Kaufmann GF, Kirchdoerfer RN, Mee JM, Janda KD, Wilson IA. 2007. Crystal structures of a quorum-quenching antibody. J. Mol. Biol. 368:1392–1402. http://dx.doi.org/10.1016/j.jmb.2007.02.081. 35. Li Y, Kilpatrick J, Whitelam GC. 2000. Sheep monoclonal antibody fragments generated using a phage display system. J. Immunol. Methods 236:133–146. http://dx.doi.org/10.1016/S0022-1759(99)00227-6. 36. Miyairi S, Tateda K, Fuse ET, Ueda C, Saito H, Takabatake T, Ishii Y, Horikawa M, Ishiguro M, Standiford TJ, Yamaguchi K. 2006. Immunization with 3-oxododecanoyl-L-homoserine lactone-protein conjugate protects mice from lethal Pseudomonas aeruginosa lung infection. J. Med. Microbiol. 55:1381–1387. http://dx.doi.org/10.1099/jmm.0 .46658-0.

aem.asm.org 469

Downloaded from http://aem.asm.org/ on January 10, 2014 by UNIVERSITAETS- UND LANDESBIBLIOTHEK DUESSELDORF

20.

chains. Nucleic Acids Res. 19:4133– 4137. http://dx.doi.org/10.1093/nar /19.15.4133. Charlton K, Harris WJ, Porter AJ. 2001. The isolation of super-sensitive anti-hapten antibodies from combinatorial antibody libraries derived from sheep. Biosens. Bioelectron. 16:639 – 646. http://dx.doi.org/10.1016 /S0956-5663(01)00192-0. Chusainow J, Yang YS, Yeo JH, Toh PC, Asvadi P, Wong NS, Yap MG. 2009. A study of monoclonal antibody-producing CHO cell lines: what makes a stable high producer? Biotechnol. Bioeng. 102:1182–1196. http: //dx.doi.org/10.1002/bit.22158. McGregor DP, Molloy PE, Cunningham C, Harris WJ. 1994. Spontaneous assembly of bivalent single chain antibody fragments in Escherichia coli. Mol. Immunol. 31:219 –226. http://dx.doi.org/10.1016/0161-5890 (94)90002-7. Strachan G, Grant SD, Learmonth D, Longstaff M, Porter AJ, Harris WJ. 1998. Binding characteristics of anti-atrazine monoclonal antibodies and their fragments synthesised in bacteria and plants. Biosens. Bioelectron. 13:665– 673. http://dx.doi.org/10.1016/S0956-5663(98)00022-0. Pearson JP, Pesci EC, Iglewski BH. 1997. Roles of Pseudomonas aeruginosa las and rhl quorum-sensing systems in control of elastase and rhamnolipid biosynthesis genes. J. Bacteriol. 179:5756 –5767. Papaioannou E, Wahjudi M, Nadal-Jimenez P, Koch G, Setroikromo R, Quax WJ. 2009. Quorum-quenching acylase reduces the virulence of Pseudomonas aeruginosa in a Caenorhabditis elegans infection model. Antimicrob. Agents Chemother. 53:4891– 4897. http://dx.doi.org/10.1128 /AAC.00380-09. Charlton KA, Porter AJ. 2002. Isolation of anti-hapten specific antibody fragments from combinatorial libraries. Methods Mol. Biol. 178:159 –171. Hayhurst A, Harris WJ. 1999. Escherichia coli skp chaperone coexpression improves solubility and phage display of single-chain antibody fragments. Protein Expr. Purif. 15:336 –343. http://dx.doi.org/10.1006/prep .1999.1035. Tan MW, Mahajan-Miklos S, Ausubel FM. 1999. Killing of Caenorhabditis elegans by Pseudomonas aeruginosa used to model mammalian bacterial pathogenesis. Proc. Natl. Acad. Sci. U. S. A. 96:715–720. http://dx .doi.org/10.1073/pnas.96.2.715.