Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, March 2009, p. 918-925, Vol. 53, No. 3
0066-4804/09/$08.00+0 doi:10.1128/AAC.00766-08
Copyright © 2009, American Society for Microbiology. All Rights Reserved.

Targanta Therapeutics Incorporated, 7170 Frederick Banting, St. Laurent, Quebec H4S 2A1, Canada,1 MicroTEM Inc., P.O. Box 1107, 101 Chalmers St., Elora, Ontario N0B 1S0, Canada,2 Guelph Regional Integrated Imaging Facility, New Science Complex, 488 Gordon St., University of Guelph, Guelph, Ontario N1G 2W1, Canada,3 Department of Molecular and Cellular Biology, New Science Complex, 488 Gordon St., University of Guelph, Guelph, Ontario N1G 2W1, Canada4
Received 12 June 2008/ Returned for modification 19 October 2008/ Accepted 11 December 2008
|
|
|---|
|
|
|---|
Oritavancin is a semisynthetic lipoglycopeptide in clinical development for the treatment of serious gram-positive infections. It exerts activity against both susceptible and methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococci. The rapidity of its bactericidal activity against exponentially growing S. aureus cells (
3-log reduction within 15 min to 2 h against methicillin-sensitive S. aureus [MSSA], MRSA, and vancomycin-resistant S. aureus [VRSA]) is one feature that distinguishes it from the prototypic glycopeptide vancomycin (31). Recent work demonstrated that oritavancin has multiple mechanisms of action that can contribute to the cell death of exponentially growing S. aureus cells, including the inhibition of cell wall synthesis by both substrate-dependent and -independent mechanisms (2, 4, 48), disruption of membrane potential and increasing membrane permeability (32), and inhibition of RNA synthesis (4). The ability of oritavancin but not vancomycin to interact with the cell membrane, leading to a loss of membrane integrity and collapse of transmembrane potential, correlates with the rapidity of oritavancin bactericidal activity (32). Mechanisms of action beyond substrate-dependent cell wall synthesis inhibition have not been described to date for vancomycin; consequently, vancomycin typically requires 24 h and actively dividing cells to exert bactericidal activity (7, 31). With this in mind, we sought to characterize oritavancin activity in vitro against S. aureus in slow-growing and biofilm states.
(Part of this work was presented at the 47th Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, IL, 17 to 20 September 2007 [7].)
|
|
|---|
Nutrient-depleted CAMHB. Nutrient-depleted CAMHB (29) from each respective strain was prepared by three rounds of inoculation of CAMHB with exponential-phase bacteria, incubation overnight at 37°C with rotation (225 rpm), and centrifugation (8,000 x g for 30 min) to remove bacteria. After the final round of inoculation, growth, and centrifugation, the pH of the nutrient-depleted CAMHB was adjusted to pH 7.0, and the spent medium was filter sterilized using a 0.22-µm membrane (Corning Incorporated, Corning, NY).
Antibacterial agents and concentrations. Antibacterial agents were tested at pharmacologically achievable concentrations that have been determined from clinical studies: concentrations were chosen to approximate free peak (fCmax) and free trough levels in plasma following administration of standard doses in humans. For oritavancin, fCmax and free trough levels from a standard dose of 200 mg (47) as well as an additional concentration that approximates the fCmax following a single 800-mg dose (fCmax800) in humans were used (18). Oritavancin diphosphate powder (Targanta Therapeutics, Cambridge, MA) was dissolved in water containing 0.002% (vol/vol) polysorbate 80; polysorbate 80 was also maintained at 0.002% in assays to minimize oritavancin loss to the surface of vessels (3), except where indicated. Concentrations that approximate the fCmax and free trough levels in plasma when administered at standard dosages for the prototypical glycopeptide vancomycin, the oxazolidinone linezolid, and the lipopeptide daptomycin were determined from pharmacokinetic data and protein binding values reported in their respective package inserts (vancomycin, Vancocin; linezolid, Zyvox; daptomycin, Cubicin). The approximation of the rifampin fCmax was derived from data reported previously by Burman et al. (8).
Time-kill studies.
Nutrient-depleted CAMHB containing diluted antimicrobial agents was inoculated with stationary-phase bacteria from cultures of S. aureus strains grown overnight at approximately 107 CFU/ml. Other experiments compared the killing of stationary- and exponential-phase S. aureus ATCC 29213 cells when inoculated into nutrient-depleted CAMHB containing the test agents. For assays involving daptomycin, nutrient-depleted CAMHB was supplemented with 50 µg/ml CaCl2 (11). All time-kill studies were performed using 96-well deep-well plates at 37°C with rotation (225 rpm) in a total volume of 750 µl. To prevent drug carryover during serial dilution plating, aliquots of the drug-challenged culture were added to an equal volume of activated charcoal suspension (25 mg/ml) (6). Bactericidal activities of the antimicrobial agents were defined as a reduction in viable cell counts of
3 log at 24 h relative to cell counts in the starting inoculum (35). Experiments were repeated at least three times and produced similar results; results from one experiment are presented.
Short-duration (2-h) time-kill studies were performed using membrane assay buffer (see below) to characterize the killing of S. aureus ATCC 29213 cells under conditions used in the membrane depolarization and permeability assays. Exponential- and stationary-phase S. aureus ATCC 29213 cells were diluted to an optical density at 600 nm [OD600] of 0.005 (approximately 106 CFU/ml) in membrane assay buffer (10 mM HEPES-Cl [pH 7.5], 50 µg/ml CaCl2) with and without 5 mM glucose, respectively (glucose was included or omitted to prevent the de-energization or energization of exponential- or stationary-phase cell membranes, respectively). Experiments were initiated by the addition of antimicrobial agents at the indicated concentrations, and bacteria were enumerated by serial dilution plating. Testing with oritavancin was done in the absence of 0.002% polysorbate 80 to reflect conditions used in membrane depolarization and permeability assays. Experiments were repeated three times and produced similar results; results from one assay are presented.
Measurements of membrane depolarization and permeability.
Membrane depolarization was monitored using the fluorescent probe 3,3'-dipropylthiadicarbocyanine iodide [DiSC3(5)] (Invitrogen Corporation, Carlsbad, CA), which partitions into the plasma membrane in proportion to the membrane potential. Dissipation of the membrane potential releases the probe, leading to an increase in fluorescence. Previous studies with the glycopeptide telavancin (19) and the lipopeptide daptomycin (42) have used this probe to demonstrate the membrane-perturbing activity of these drugs against exponential-phase cells. S. aureus ATCC 29213 was chosen for testing in membrane studies. Bacteria were grown overnight in CAMHB and subcultured the following day in CAMHB to exponential phase (OD600
0.25). Exponential- and stationary-phase cells were washed in membrane assay buffer with and without 5 mM glucose, respectively, and resuspended at an OD600 of 0.25. DiSC3(5) was added to a final concentration of 1.5 µM, and the solution was incubated in the dark at ambient temperature for 30 min to allow the loading of the fluorescent dye into cell membranes. After the loading period, cells were diluted 50-fold (OD600 of 0.005) in depolarization buffer with or without glucose for exponential- or stationary-phase cells, respectively. Assays were initiated by the addition of antimicrobial agents over a range of concentrations and were monitored in real time by fluorescence spectroscopy (excitation wavelength of 612 nm and emission wavelength of 665 nm) for a period of 30 min. Note that 0.002% polysorbate 80 was found to interfere with fluorescence in these assays and was therefore omitted from the assay. Experiments were repeated three times and produced similar results; results from one assay are presented.
Changes in bacterial membrane permeability were quantified using the fluorescent dye pair Syto 9 and propidium iodide: bacterial membrane damage (increased permeability) allows the otherwise membrane-impermeable dye propidium iodide to enter the cell and displace the permeative dye Syto 9, leading to a loss of fluorescence. Bacteria were prepared as described above for the membrane depolarization assay, but Syto 9 and propidium iodide (Invitrogen Corporation) were added at 5 µM and 30 µM, respectively (32). Fluorescence spectroscopy (excitation wavelength of 485 nm and emission wavelength of 535 nm) was monitored for 30 min following the addition of antimicrobial agents. As mentioned above, 0.002% polysorbate 80 was omitted from the assay to prevent interference with fluorescence determinations. Experiments were repeated three times and produced similar results; results from one assay are presented.
Determination of ultrastructural effects of oritavancin and vancomycin on stationary-phase cells by transmission electron microscopy. Stationary-phase MRSA ATCC 43300 cells (5 x 107 CFU/ml) were exposed to 1 µg/ml oritavancin (2x its broth microdilution MIC in the absence of 0.002% polysorbate 80) or 16 µg/ml vancomycin (16x its broth microdilution MIC) in nutrient-depleted CAMHB for 3 h. Bacteria were fixed in 2.5% glutaraldehyde to cross-link proteins and help preserve morphological structure. Prior to embedding, the samples were treated with fresh 2.5% (vol/vol) glutaraldehyde in HEPES buffer (pH 6.8) for 2 h. The samples were then postfixed in 2.0% (wt/vol) osmium tetroxide, followed by en bloc staining with 2.0% (wt/vol) uranyl acetate, as a heavy-metal stain, to add contrast to the cells. The cells were then dehydrated through a series of ethanol washes and then embedded in LR White resin. Once polymerized by curing, each culture sample was thin sectioned and stained by uranyl acetate and lead citrate to view the internal cellular constituents and the juxtaposition of cell envelope layers such as the plasma membrane and cell wall. Transmission electron microscopy was used to view the thin sections using a Philips CM10 microscope under standard operating conditions at 100 kV.
Determination of MBEC. In vitro biofilms were established using the MBEC Physiology & Genetics Assay plate (Innovotech; Edmonton, AB, Canada) according to the manufacturer's protocol (20). The MBEC system also allows the determination of the MIC of the test agent for planktonic cells shed from the biofilm as well as the minimum biofilm eradication concentration (MBEC), the concentration of antimicrobial agent required to sterilize the biofilm after 24 h of exposure. Briefly, 150 µl of bacterial inocula at 107 CFU/ml in tryptic soy broth was aliquoted into each well of an MBEC plate. Biofilms were established on the MBEC peg lid for 24 h in a rotary incubator at 37°C and 150 rpm. For experiments involving 72-h biofilms, MBEC peg lids were transferred each day into 96-well plates containing 150 µl/well of fresh tryptic soy broth and incubated another 24 h. MBEC peg lids with established biofilms were washed once in sterile saline (200 µl/well) and then placed onto plates containing antimicrobial agents diluted in CAMHB (200 µl/well). Antimicrobial agents were serially diluted in CAMHB in 96-well plates, and MBEC peg lids were exposed for 24 h or for the indicated times. Note that 0.002% polysorbate 80 was found to adversely affect biofilm cell numbers and was therefore omitted from MBEC determinations for oritavancin. Following antimicrobial challenge and determination of the planktonic MICs, MBEC peg lids were washed once in sterile saline and then placed into recovery plates containing CAMHB (200 µl/well). The recovery plates were sonicated for 5 min in an ultrasonic sonicating bath (VWR Aquasonic model 550D) at the maximum setting and then incubated for 24 h, and the MBECs were recorded. MBECs were determined from at least three independent experiments; results represent the ranges of MBECs obtained. To enumerate the biofilm CFU on individual control pegs, pegs were broken off the MBEC peg lid using sterile forceps placed into 1 ml of sterile saline, sonicated for 5 min, and vortexed for 1 min at the highest setting. Bacteria were then enumerated by serial dilution plating. CFU/peg counts were determined from at least three independent experiments; results presented are the averages ± standard deviations.
|
|
|---|
![]() View larger version (11K): [in a new window] |
FIG. 1. Time-kill kinetics of stationary- and exponential-phase MSSA ATCC 29213 in nutrient-depleted CAMHB. Viability was enumerated at the indicated time points by serial dilution plating. Each point represents the mean of duplicate determinations. The limit of detection is indicated as a dashed line. (A) Stationary-phase inocula challenged with estimated free trough, fCmax, fCmax800 of oritavancin, and fCmax of comparators. (B) Exponential-phase inocula challenged with estimated free trough, fCmax, and fCmax800 of oritavancin and comparators. *, untreated control; , 4 µg/ml oritavancin; , 16 µg/ml oritavancin; , 16 µg/ml vancomycin; , 4 µg/ml daptomycin; , 8 µg/ml linezolid; , 2 µg/ml rifampin.
|
![]() View larger version (10K): [in a new window] |
FIG. 2. Time-kill kinetics of stationary-phase MRSA ATCC 33591 and VRSA VRS5 by oritavancin and comparators in nutrient-depleted CAMHB. Viability was enumerated at the indicated time points by serial dilution plating. Each point represents the mean of duplicate determinations. The limit of detection is indicated as a dashed line. (A) MRSA ATCC 33591 challenged with estimated free trough, fCmax and fCmax800 of oritavancin, and fCmax of comparators. (B) VRSA VRS5 challenged with estimated free trough, fCmax and fCmax800 of oritavancin, and fCmax of comparators. *, growth control; , 4 µg/ml oritavancin; , 16 µg/ml oritavancin; , 16 µg/ml vancomycin; , 4 µg/ml daptomycin; , 8 µg/ml linezolid; , 2 µg/ml rifampin.
|
![]() View larger version (25K): [in a new window] |
FIG. 3. Measurement of oritavancin effects on membrane depolarization, permeability, and killing of MSSA ATCC 29213. (A and B) Membrane depolarization was monitored by measuring DiSC3(5) fluorescence. (C and D) Permeabilization of the cell membranes was monitored by measuring Syto 9 and propidium iodide fluorescence. Note that in D, the curve for vancomycin versus stationary-phase cells overlaps the curve for daptomycin versus stationary-phase cells. RFU, relative fluorescence units. (E and F) Killing kinetics of stationary- and exponential-phase inocula in membrane assay buffer. Glucose was omitted from the membrane assay buffer for stationary-phase cells and included at 5 mM for exponential-phase cells. The limit of detection is indicated as a dashed line. For A, C, and E, symbols are as follows: , 4 µg/ml oritavancin versus exponential-phase cells; , 4 µg/ml oritavancin versus stationary-phase cells; , untreated exponential-phase cells; , untreated stationary-phase cells. For B, D, and F, symbols are as follows: , 16 µg/ml oritavancin versus stationary-phase cells; , 16 µg/ml oritavancin versus exponential-phase cells; , 16 µg/ml vancomycin versus stationary-phase cells; , 16 µg/ml vancomycin versus exponential-phase cells; , 4 µg/ml daptomycin versus stationary-phase cells; , 4 µg/ml daptomycin versus exponential-phase cells.
|
Time-kill studies using membrane assay buffer over a short duration of exposure also showed that the rate of killing of stationary-phase MSSA by oritavancin was decreased compared to that of the exponential-phase inoculum (Fig. 3E). The rapid bactericidal activity of oritavancin against exponential-phase cells was exemplified by a 3.2-log reduction in CFU within 15 min when tested at 16 µg/ml, its predicted fCmax800. Bacterial killing was also seen with the fCmax of oritavancin (4 µg/ml; 2.9-log reduction) and daptomycin (4 µg/ml; 3.4-log reduction) within 2 h of exposure (Fig. 3E and F). In contrast, stationary-phase cells exhibited approximately 1.5-log and 0.9-log reductions in CFU within the 2-h time period following exposure to oritavancin at the fCmax800 and fCmax, respectively. Daptomycin activity was similarly reduced, as it exerted a 0.7-log reduction in CFU at its fCmax (Fig. 3F). Vancomycin did not effect any change on bacterial counts of either inoculum over the short exposure time of the assay (Fig. 3F).
Oritavancin targets the septum of stationary-phase MRSA ATCC 43300 cells. We recently examined the effect of oritavancin on the ultrastructure of exponential-phase MRSA ATCC 43300 cells by transmission electron microscopy and observed septal deformations and a loss of staining of the nascent septal cross wall, the "midline" (30), in exposed cells (5). These effects were not seen following vancomycin exposure. In this study, qualitative differences were evident upon examination of the stationary-phase culture compared to exponential-phase cells: cell ghosts were present but at a low frequency (data not shown), and septa appeared broader (Fig. 4A) than in exponential-phase cells (Fig. 4B). Furthermore, an electron-dense material was present throughout the extracellular space and attached to the surface of stationary-phase cells (Fig. 4A, C, and D). Septa of oritavancin-treated cells were also broad, but staining of the midline was conspicuously absent (Fig. 4C), corroborating observations of exponential-phase cells (5). The midline was evident in vancomycin-treated cells (Fig. 4D), which overall had an ultrastructural appearance similar to that of the untreated cells.
![]() View larger version (152K): [in a new window] |
FIG. 4. Ultrastructural analysis of stationary-phase MRSA ATCC 43300 by transmission electron microscopy of thin sections. (A) Untreated control cells. The arrow indicates the septal midline. (B) Exponential-phase MRSA ATCC 43300 is shown for comparison. The septum is not as broad as in stationary-phase cells (compare to A). The arrow indicates the septal midline. (C) Cells exposed to 1 µg/ml oritavancin for 3 h. Note the absence of a well-defined midline. (D) Cells exposed to 16 µg/ml vancomycin for 3 h. The arrow indicates the septal midline.
|
The capacity of each strain to form a biofilm on the pegs of the MBEC plate was determined by enumerating the CFU attached to the peg surface (CFU/peg). Biofilm cell densities on the pegs for each strain were (2.9 ± 2.4) x 105 CFU/peg for MSSA ATCC 29213, (2.2 ± 1.7) x 105 CFU/peg for MRSA ATCC 33591, and (2.6 ± 1.1) x 105 CFU/peg for VRSA VRS5 after 24 h of incubation. Planktonic MICs determined for comparator antimicrobial agents in the MBEC assay were within the CLSI quality control ranges (Table 1). Although the testing methodologies were not identical to the conventional broth microdilution MIC method based on CLSI guidelines (10), oritavancin planktonic MICs were also within the quality control range (0.5 to 2 µg/ml) for MSSA ATCC 29213 as determined in the absence of 0.002% polysorbate 80 (13). The growth of S. aureus in a biofilm resulted in dramatic decreases in the antimicrobial activities of vancomycin and linezolid as measured by the concentration of antimicrobial agent needed to sterilize the 24-hour biofilm (MBEC) compared to their respective planktonic MICs (Table 1): MBECs for both agents were >128 µg/ml against all three strains. In contrast, oritavancin MBECs ranged from 2 to 8 µg/ml against the S. aureus strains (Table 1) and were within 1 doubling dilution of their respective planktonic MICs in each experiment.
|
View this table: [in a new window] |
TABLE 1. Oritavancin exhibits antibiofilm activity in vitro against S. aureus strains of different resistance phenotypesa
|
|
|
|---|
A reduced rate of killing of stationary-phase cells by oritavancin was observed in either nutrient-depleted CAMHB or membrane assay buffer and was concomitant with a reduced rate of depolarization and permeabilization of cell membranes. Although we cannot exclude the possibility that oritavancin targets are less abundant in stationary-phase cells, these findings are likely an indication that the bactericidal mechanism of action of oritavancin is linked to the energized state of the cell: membrane potential is known to be greater in exponential-phase S. aureus cells (–167 mV) than in stationary-phase cells (–123 mV) (23). Membrane potential and the protonated state of the cell wall are thought to regulate cell wall hydrolase (autolysin) activity (24), and oritavancin-induced changes in the membrane potential of stationary-phase cells may account for the observed loss of the septal midline: the decreased staining intensity of the septal midline in stationary-phase S. aureus cells may have reasonably resulted from a loss of the chemically reactive sites that are exposed (and that bind the heavy metal stain) when cell wall hydrolases cleave septal polymers of the nascent cross walls during division (30, 46). Such a loss of the septal midline in stationary-phase cells corroborates previously reported observations of exponential-phase cells (5). Another possibility is that the inhibition of nascent cross-wall synthesis could also account for the loss of the midline: oritavancin inhibits cell wall synthesis (4) and can directly inhibit the transglycosylase activity of S. aureus penicillin-binding protein 2 (48), an essential transglycosylase-transpeptidase that is localized to the division septum during cell division (36). Thus, while a full understanding of the means by which oritavancin causes cell death requires further investigation, we show in the current work that it likely involves a common mechanism against stationary- and exponential-phase cells.
The potency of oritavancin against S. aureus in vitro biofilms is highlighted by (i) MBEC values within 1 doubling dilution of their planktonic MICs, (ii) sterilization of MSSA biofilms within 1 h, and (iii) sterilization of 72-h biofilms of MSSA that also had increased cellular densities at MBECs within 1 doubling dilution of its planktonic MIC against the test strain. As was seen against stationary-phase cells, vancomycin and linezolid did not sterilize the biofilms, confirming previously reported findings of greatly reduced activities of these agents in in vitro biofilm models (9, 17, 21, 39). Thus, in the in vitro biofilms described here, a tolerant population of cells was present. The finding that the stratification of DNA synthesis activity in an in vitro colony biofilm model in which the vast majority of cells synthesizing DNA (and, therefore, actively dividing) was at the air interface (37) indicates that biofilms are composed of a mixed population of cells with different levels of metabolic activity. Furthermore, another study reported that only a subpopulation of cells in a biofilm were readily killed by the fluoroquinolone ofloxacin (43). Regardless of whether the in vitro (24- or 72-h) biofilms of S. aureus described here were composed of cells of different metabolic activities, oritavancin sterilized the biofilms at a concentration that was 1 doubling dilution higher than that needed to kill planktonic cells. Although rifampin exhibited reduced activity against the in vitro S. aureus biofilms, its MBEC (4 µg/ml) approached pharmacologically achievable concentrations (fCmax
2 µg/ml) (8). However, development of rifampin resistance has been observed when this agent is used alone in in vitro biofilm models (32) and in vivo (22, 28), which therefore should restrict its use to combination therapy.
The development of antimicrobial agents and therapies that are active against biofilms and tolerant cells would benefit the treatment of infections that harbor cells in these states (14). Unfortunately, these types of infections are often complicated by ischemia (i.e., sequestra of osteomyelitis and chronic wound of diabetic foot ulcers), compromising antimicrobial exposure and activity at the infection site. Thus, new agents and therapies must overcome significant hurdles. The present study shows that oritavancin exerts bactericidal activity against stationary-phase S. aureus cells, likely by its capacity to disrupt the membrane integrity of susceptible bacteria. Furthermore, its potency against the in vitro biofilm was remarkable in light of the observed significant tolerance of the biofilm to comparator antimicrobial agents. To date, oritavancin has shown efficacy in infections that likely harbor tolerant cells, including rat central venous catheter (38), rat granuloma pouch (26), and rabbit endocarditis (25, 40) models. Studies to examine the efficacy of oritavancin in in vivo models of biofilm infections are warranted to confirm its promising activity against cells in a tolerant state in vitro.
With the exception of R.H and T.B., all authors of the manuscript are Targanta Therapeutics employees.
Published ahead of print on 22 December 2008. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»