![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Previous Article | Next Article 
Antimicrobial Agents and Chemotherapy, May 2004, p. 1461-1468, Vol. 48, No. 5
0066-4804/04/$08.00+0 DOI: 10.1128/AAC.48.5.1461-1468.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Microbiology Unit, Division of Infection and Immunity, Eastman Dental Institute for Oral Health Care Sciences, University College London, London, United Kingdom
Received 21 October 2003/ Returned for modification 7 November 2003/ Accepted 8 January 2004
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Dental plaque is a biofilm which forms on the nonshedding surfaces of the oral cavity (33). If left untreated, the succession of dental plaque development can lead to serious complications, such as caries, gingivitis, and periodontitis. A typical plaque removal regimen might involve brushing with either a manual or an electric toothbrush, followed by the use of a proprietary antimicrobial mouthwash for 30 s. The latter products are commonly based on a dilute alcohol solution containing an antimicrobial compound, such as chlorhexidine gluconate (CHX), although many other formulations are available (28). CHX is also the active ingredient in many commercially available disinfectants and antiseptics. CHX (C22H30Cl2N102C6H12O7) is a cationic bisbiguanide with a molecular mass of 898 Da. It possesses broad antibacterial activity in combination with low mammalian toxicity and the ability to bind to skin and mucous membranes. The CHX molecule reacts with negatively charged groups on the cell surface, causing an irreversible loss of cytoplasmic constituents, membrane damage, and enzyme inhibition. At high concentrations, CHX results in extensive cell damage, coagulation of cytoplasm, and precipitation of proteins and nucleic acids. A CHX concentration of 0.2% is deemed to be the most effective as a mouthwash and as such is considered the "gold standard" (15).
It has been shown in an in vitro model that CHX exposure times on the order of 30 s have very little effect on the number of viable bacteria which can be recovered from oral biofilms. In particular, 0.2% CHX has been shown to be ineffective against dental plaque in an in vitro model after 5 min of exposure, requiring 60 min of exposure to achieve 2-log10 to 5-log10 killing (24).
Confocal laser scanning microscopy (CLSM) uses a tightly focused beam of laser light and a pinhole aperture to capture a series of thin optical sections, termed an image stack, through a (biological) sample. Analysis of CLSM images has been used in conjunction with vital staining techniques to determine the architecture and spatial distribution of vitality for organisms within oral biofilm (1, 21, 35). These techniques have been refined by treating viable and nonviable bacteria as distinct populations and then comparing the spatial distribution trends between these two cohorts to produce viability profiles (9). These processes utilize a function of image fluorescence versus the depth of the optical section in the confocal image stack; viability profiles typically show the progression from viable fluorescence to nonviable fluorescence with increasing depth in the biofilm. This phenomenon is also apparent in the horizontal plane; the outer layers of the biofilm surface features exhibit more viable fluorescence than do the inner regions, which exhibit nonviable fluorescence (12, 22).
BacLight LIVE/DEAD stain (Molecular Probes, Eugene, Oreg.) is a two-component system utilizing SYTO9 (proprietary nomenclature) and propidium iodide to differentiate between viable and nonviable bacterial cells, respectively. Bacteria with damaged membranes are stained as nonviable, whereas bacteria with intact membranes are deemed viable. While these guidelines may not hold true for every bacterium in every circumstance (26), the fidelity of the system is sufficient to elucidate depth-related trends in cell vitality within multispecies populations of bacteria comprising a biofilm. Since CHX damages bacterial membranes, its bactericidal action can be measured in real time with BacLight LIVE/DEAD stain.
We used CLSM in conjunction with fluorescent indicators of membrane integrity to monitor, in real time, the effects of CHX on oral biofilms in terms of changes in their viability profiles.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The biofilms reached a steady state after 7 days, and the disks were removed from the CDFF at 17 to 27 days after inoculation. Hence, biofilms at steady state (but at different ages) were investigated to ascertain whether any generalizations could be made with regard to the effects of CHX on viability profiles within such biofilms.
CLSM and application of CHX. For CLSM, two solutions were prepared. A 10% CHX solution was prepared from a 20% stock (Sigma-Aldrich Co. Ltd., Gillingham, United Kingdom). A viewing solution comprising 8 ml of distilled water (dH2O) containing 2 µl each of components A and B of BacLight LIVE/DEAD stain was prepared.
A hydroxyapatite disk containing a biofilm was dipped in 1 ml of saline to remove excess medium and unattached cells. The disk then was placed, biofilm up, into a small cell culture dish (Bibby Sterilin Ltd., Stone, United Kingdom), and the viewing solution was carefully poured into the dish. This volume of liquid submerged the entire disk and biofilm, which then was allowed to develop in the dark for 10 min. Biofilms were examined by using a fixed-stage microscope (DMLFS; Leica Microsystems, Milton Keynes, United Kingdom) incorporating a Leica TCS SP2 laser scan head mounted on a vibration-free platform. The objective lenses used were x20 and x40 water immersion dipping lenses (Leica). The regions within each biofilm selected for analysis were those which contained stacks (or "towers") that could be isolated within the confocal field of view.
A preliminary scan of a random area of a biofilm was taken to determine the optimum photomultiplier (PMT) and z-axis settings (upper and lower positions of the image stack). It was later found that the PMT gain and offset values were best set so as to yield a slightly wider range than was dictated by the preliminary scan. While this strategy reduced image quality in terms of the range of pixel brightness information, it allowed the quantification of fluorescence values above and below the levels of the original image. Shifting the PMT settings in this manner helped to minimize saturation.
The CLSM control software was set to take a series of time-lapse scans (xyzt) at intervals of either 30 or 60 s, depending on the thickness of the image stack. Scans were taken in 8 bits at a resolution of 512 by 512 pixels. This relatively low image quality was used in order to minimize the scan time, an estimate of which was displayed by the control software. The z-axis step size was the major determinant of the scan time, and this was the value which was manipulated in order to balance image quality with time-point resolution. The length of the time-lapse series was typically 10 to 15 min.
Immediately after completion of the first scan of the time-lapse series, 160 µl of the 10% CHX solution was added to the cell culture dish to yield a final concentration of CHX in the 8-ml viewing solution of 0.2% (wt/vol). The CHX solution was injected via a pipette (Eppendorf AG, Hamburg, Germany) at an angle perpendicular to the disk with the aim of maximizing mixing while avoiding direct disruption of the biofilm by fluid shear. Similar experiments in which 40 µl of the 10% CHX solution was added to the viewing solution to achieve a final concentration of 0.05% (wt/vol) CHX were conducted. The resulting collections of confocal sections were archived onto optical disks as time-depth-PMT channel-coded image files (in tagged image file format [TIFF]) by using Leica TCS NT software.
A series of complementary time-lapse control experiments in which CHX was not added to the viewing solution also were undertaken.
Image analysis. Image stacks were analyzed by using the Java-based image analysis program ImageJ (8; W. Rasband, ImageJ 1.29x [PC-Windows] freeware, http://rsb.info.nih.gov/ij/). The individual time-depth-PMT channel-coded TIFF image files were first reassembled into usable TIFF stacks with a specially written ImageJ plug-in. The relevant ImageJ commands are given in parentheses below.
Projection images were assembled for each time point and fluorescence channel to produce a single image based on the sum of pixel brightness values through the image stack (ImageJ: z-project). The projections were converted from 32 bits back to 8 bits by using a scaled method, and the resulting images were formed into an image stack. This time-lapse stack then was analyzed to divulge time-related trends in total image fluorescence for both the viable and the nonviable channels (ImageJ: plot z-axis profile). Sagittal (elevation view) time-lapse projections were constructed by similar methods.
Fluorescence intensity profiles along the z axis of the original image stacks were determined for both the viable and the nonviable fluorescence channels (ImageJ: plot z-axis profile). These raw data then were exported to a spreadsheet program (Microsoft Excel), and the process was repeated for each point in the time-lapse series. Once the z-axis profiles were in spreadsheet format, the data were normalized; i.e., the maximum fluorescence value for a particular channel at a particular time point was allocated a value of 1 unit. Viability profiles then were constructed by subtracting the normalized nonviable z-axis profile values from the corresponding viable values and plotting the data against the depth of the optical section in the biofilm (9).
The rates of biofilm contraction, as revealed by the sagittal projections, were measured by object tracking. This involved noting the z-axis positions of distinguishing features within the image stack, such as the top of the biofilm or a particularly bright microcolony, and tracking the movement of these features over time. The corresponding increase in image brightness due to increasing fluorescence per unit volume was calculated and factored into the sum of projected fluorescence values.
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Preliminary experiments showed that CHX exposure caused optical section fluorescence within the biofilm to increase, often resulting in saturation of the captured images, i.e., too many pixels at the maximum brightness value (data not shown). This phenomenon was attributed to the observed contraction of the biofilm causing a corresponding increase in the fluorescence per unit volume (i.e., within the thickness of an optical section).
Upon exposure to 0.2% CHX, time-lapse projection images showed evidence of changes in oral biofilm 1 vitality and structure (Fig. 1). Sagittal projections (Fig. 2) showed that 0.2% CHX caused biofilm 1 to contract at a rate of 1.176 µm min–1, which was measured by object tracking over a period of 15 min (Fig. 3). The influence of sample contraction on increasing image fluorescence per optical section was calculated with the equation given in the legend to Fig. 3.
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Examination of biofilms while they were immersed in a viewing solution allowed the diffusion of aqueous solutes into the biofilms to be studied in the "natural" hydrated state. This detail is important, as the major component of any biofilm is water. Dehydration and/or fixing in resin disrupts the structural motifs and spatial distributions of bacteria within a biofilm. Phosphate-buffered saline, which would normally be used to simulate physiological conditions, could not be incorporated into the viewing solution because CHX reacts with phosphate and chloride radicals, resulting in double decomposition and the slow crystallization of insoluble salts. Control experiments showed that exposing the biofilms to the dH2O-based viewing solution had no adverse osmotic effects on the biofilms with regard to structure or viability (Fig. 5b).
An exposure time of 30 s is most often quoted to describe typical mouthwash use, although poor user compliance may reduce the exposure time still further. A recent study suggested that there was no significant difference in the plaque index between patients using a 30-s rinse with a 0.12% CHX solution and those using a 60-s rinse with a 0.2% CHX solution (16). CHX has been shown to persist in the mouth by binding to mucosal surfaces and also to the pellicle and saliva. At 2 h after rinsing with 0.2% CHX, saliva retains antibacterial properties (25) and suppresses counts of bacteria in saliva for over 12 h (27). This phenomenon has been attributed to the gradual desorption of CHX into the mouth, creating a bacteriostatic milieu (13); however, it is more likely that the antimicrobial effect is due to tooth-bound CHX (15).
A study of the mass transport of fluorescence-labeled dextrans of different sizes (dextran with a molecular weight of 3,000 [3K-Dex] to 240K-Dex) into oral biofilms (32) showed that biofilms did indeed inhibit the diffusion of macromolecules and that the extent of the inhibition was greater than that which was predicted. This inhibition of diffusion was attributed to the tortuosity of the biofilm matrix, i.e., the extremely convoluted diffusion paths which the molecules must traverse. A model was produced for the diffusion of dextrans above 10K-Dex; however, this model did not encompass the diffusion of 3K-Dex. This was ascribed to the EPS possessing "pore" diameters wider than 2.6 nm but narrower than 4.6 nm. These pore diameters were based on the predicted diameters of the dextrans, and since the molecular mass of CHX (897.77 Da) is well below that of 3K-Dex, it is unlikely that the diffusion of CHX into biofilms is significantly retarded by a molecular sieving effect. It is more likely that ionic interactions occur between the positively charged CHX molecules and the negatively charged extracellular matrix. These ionic interactions are understood to reduce the diffusion coefficients of fluorescent probes within biofilms of Lactococcus lactis and Stenotrophomonas maltophilia about 50-fold (7). Conversely, for the diffusion of aminoglycosides, such as tobramycin, into biofilms of Pseudomonas aeruginosa, it was determined that the interactions between the positively charged molecules and the negatively charged matrix did not constitute a major mechanism of biofilm resistance (3). Mathematical modeling has also suggested that the diffusion of stoichiometrically reacting solutes through biofilms is not significantly retarded (30).
Mathematical modeling has been applied specifically to the diffusion of CHX into dental plaques (31). If the aqueous diffusion coefficient of CHX in water at 30°C (an assumption of the temperature of the solution held in the mouth) is 4.2 x 10–6 cm2 s–1, then for a dental plaque 260 µm thick, the time needed for diffusion through the biofilm to the substratum is 298 s. If one supposes that the centers of the biofilm structures examined in our study were at most 50 µm from the bulk fluid flow (Fig. 1), then the diffusion time would be on the order of 1 min. This analysis does not entirely account for the delay of approximately 5 min before CHX was observed to affect cell membrane integrity, suggesting that oral biofilms possess intrinsic resistance to CHX which goes beyond their ability to impede diffusion (6).
Exposure to CHX does not kill bacteria immediately; a delay of 20 s has been reported for Escherichia coli and P. aeruginosa (4), and a delay of 30 s has been reported for oral streptococci (23). In this study with biofilms of oral bacteria, the bactericidal effects of both 0.05 and 0.2% CHX were detectable at 3 to 5 min. This anomaly could indicate that biofilms of oral bacteria are intrinsically less susceptible to the effects of CHX than are their planktonic counterparts. It is reasonable to assume that there will be no significant delay between the cell membrane becoming compromised and the penetration of a nonviability stain into cells, since fluorophores will already be present in close proximity to cells.
The mechanisms underlying the contraction of oral biofilm upon exposure to CHX are probably related to ionic interactions between the negatively charged EPS matrix, which comprises the bulk of the volume of biofilm, and the positively charged CHX molecules. These interactions will change the physicochemical properties of the EPS: solubility, hydrophobicity, and localized charge along the polymer chains. Changes in charge will in turn affect the tertiary structure of the EPS chains and the degree of bonding with adjacent strands. As the positive CHX interacts with the negative EPS, the net charge of the matrix will shift toward neutral, reducing the repulsive forces between charged moieties, allowing closer associations to occur between polymeric strands, and reducing the volume occupied by the biofilm. A more compact matrix may further inhibit the diffusion of solutes, including CHX, into the biofilm due to the tightening of the apparent molecular sieve.
The phenomenon of biofilm contraction appears to be related to the concentration of CHX, since the contraction observed at 0.2% was distinct and quantifiable, while at 0.05% the contraction was very slight and nonuniform. Observations made during the exposure of biofilm 3 to 0.05% CHX showed a water channel which appeared to open up. Image analysis estimated that the area (i.e., the number of pixels) occupied by the water channel increased 17% over 15 min. It is possible that conformational changes in biofilm structure, such as the opening up of this water channel, could actually assist in the diffusion of CHX further into deeper layers. Biofilm exposed to CHX at 0.5% also exhibited marked contraction (data not shown due to image saturation; see Materials and Methods).
Biofilm contraction is unlikely to be due to any synergistic effect between CHX and the osmotic pressure of dH2O. Control experiments in which phosphate-buffered saline was incorporated into the viewing solution were undertaken. Although double decomposition of the salts was apparent as the formation of a milky precipitate, the biofilm was still seen to contract and shift toward nonviable fluorescence (data not shown).
Further work is necessary to understand the interplay between the delay of the action of CHX on cell viability and the apparently immediate contraction of the biofilm in order to determine the effects exerted by these phenomena on the observations described in this study and to determine whether contraction of the biofilm assists or impedes CHX penetration. Future experiments are being devised to evaluate the effects of different concentrations of CHX and other antimicrobial compounds. It would also be of great interest to use longer time frames to determine when the biofilm ceases contraction in response to CHX; however, the massive amounts of data generated are a limiting factor.
While total fluorescence measurements were capable of measuring the antimicrobial effects of CHX at a concentration of 0.2% (wt/vol) (Fig. 4), this technique was not sensitive enough to effectively elucidate the bactericidal effects of 0.05% CHX (Fig. 5). When depth was included as a variable, the time-dependent bactericidal effects of 0.05% CHX were much more easily visualized (Fig. 7) and were magnified further by the construction of viability profiles (Fig. 9). We have shown that viability profiles through oral biofilm change upon exposure to CHX and suggest that they are more sensitive at detecting the antimicrobial effects of lower concentrations of CHX and presumably other membrane-active antimicrobial compounds than are total fluorescence measurements (i.e., nonconfocal fluorescence microscopy).
These results suggest that viability profiling can be used to investigate the penetration and antimicrobial effects of existing and novel membrane-active biocidal agents on biofilm. By incorporating a depth-related function into the analysis of the image stack, the sensitivity of total fluorescence measurements was improved considerably. The effect of the contraction phenomenon on the susceptibility of the biofilm to CHX remains unclear. Whereas compaction of the EPS matrix likely would inhibit the diffusion of CHX molecules into the biofilm, the opening up of water channels would facilitate the transport of the antimicrobial agent to the inner regions of the biofilm.
|
ACKNOWLEDGMENTS |
|---|
Funding for this project was provided by Philips Oral Healthcare Inc., Snoqualmie, Wash.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Clin. Vaccine Immunol. | Clin. Microbiol. Rev. |
|---|---|
| J. Clin. Microbiol. | ALL ASM JOURNALS |
ct)/d] x f, where 
