Journal of Dental Research

Journal of Dental Research http://jdr.sagepub.com/

Architecture of Intact Natural Human Plaque Biofilms Studied by Confocal Laser Scanning Microscopy S.R. Wood, J. Kirkham, P.D. Marsh, R.C. Shore, B. Nattress and C. Robinson J DENT RES 2000 79: 21 DOI: 10.1177/00220345000790010201 The online version of this article can be found at: http://jdr.sagepub.com/content/79/1/21

Published by: http://www.sagepublications.com

On behalf of: International and American Associations for Dental Research

Additional services and information for Journal of Dental Research can be found at: Email Alerts: http://jdr.sagepub.com/cgi/alerts Subscriptions: http://jdr.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://jdr.sagepub.com/content/79/1/21.refs.html

Downloaded from jdr.sagepub.com by guest on July 13, 2011 For personal use only. No other uses without permission.

S.R. Wood'*, J. Kirkham', P.D. Marsh', R.C. Shore', B. Naltress2, and C. Robinson' Divisions of 'Oral Biology and 2Restorative Dentistry, Leeds Dental Institute, University of Leeds, Clarendon Way, Leeds LS2 9LU, United Kingdom; *corresponding author, [email protected]

J Dent Res 79(1): 21-27, 2000

Architecture of Intact Natural Human Plaque Biofilms Studied by Confocal Laser Scanning Microscopy

ABSTRACT

INTRODUCTION

Determination of the structure of human plaque will be of great benefit in the prediction of its formation and also the effects of treatment. However, a problem lies in the harvesting of undisturbed intact plaque samples from human volunteers and the viewing of the biofilms in their natural state. In this study, we used an in situ device for the in vivo generation of intact dental plaque biofilms on natural tooth surfaces in human subjects. Two devices were placed in the mouths of each of eight healthy volunteers and left to generate biofilm for 4 days. Immediately upon removal from the mouth, the intact, undisturbed biofilms were imaged by the non-invasive technique of confocal microscopy in both reflected light and fluorescence mode. Depth measurements indicated that the plaque formed in the devices was thicker round the edges at the enamel/nylon junction (range = 75-220 pm) than in the center of the devices (range = 35-215 Vim). The reflectedlight confocal images showed a heterogeneous structure in all of the plaque biofilms examined; channels and voids were clearly visible. This is in contrast to images generated previously by electron microscopy, suggesting a more compact structure. Staining of the biofilms with fluorescein in conjunction with fluorescence imaging suggested that the voids were fluid-filled. This more open architecture is consistent with recent models of biofilm structure from other habitats and has important implications for the delivery of therapeutics to desired targets within the plaque.

M icrobial biofilms consist of bacteria adhering to surfaces or interfaces and to each other, and usually embedded in an extracellular matrix (Costerton et al., 1995). They form rapidly on almost any surface that is wet and are the normal mode of colonization for micro-organisms in the environment. Biofilms of this sort can have profound implications for nature, medicine, and industry (Costerton et al., 1987). Presumably as a result of the proximity of and adhesion to the substratum, the bacteria in biofilms usually exhibit different phenotypes compared with their planktonic counterparts (Davies et al., 1993; Hoyle et al., 1993). They are also more resistant to antibacterial treatments (Brown et al., 1988), a characteristic which might be related to their slower growth rate (Nickel et al., 1985; Gristina et al., 1987), to problems of antibiotic penetration into the biofilm, or to inactivation of the agent in the biofilm (Foley and Gilbert, 1996). Architecture of the biofilm itself is likely to be an important factor in the modulation of both microbial physiology and in determining the ecology of the site. For example, microbial behavior will depend upon biofilm thickness, density, and the openness of the structure, as well as the cell:matrix ratio. Human dental plaque, which forms on tooth and soft-tissue surfaces, is a complex biofilm which is implicated in a number of oral diseases (Loesche et al., 1982; Moore et al., 1983; Loesche, 1986; Bowden, 1990). While it is relatively easy to access (for example, by mechanical harvesting or the insertion of artificial surfaces into the oral cavities of volunteers), site-specific chemical/microbial studies of the structure and organization of the biofilm have not been possible, since mechanical removal of the plaque is likely to disturb the delicate three-dimensional relationship among cells, matrix, space, and substratum. Almost all studies of biofilm structure in relation to microbial metabolism and ecology have been restricted to in vitro model systems capable of generating biofilm structures. Various culture systems such as constant-depth fermenters (Peters and Wimpenny, 1988), chemostats (Keevil et al., 1987), and flow cells (Caldwell and Lawrence, 1988) have all been used to generate both single- and mixed-species biofilms in vitro. [For a recent review of in vitro culture systems, see Sissons (1997).] Studies of biofilms produced in this way have already revealed complex structures with a heterogeneous distribution of cells, matrix, and fluid-filled voids (Caldwell et al., 1992). The significance of such heterogeneous structures in terms of natural biofilm behavior is likely to be profound. This would be especially true where human tissues are concerned, and biofilm structure/behavior may be related to pathogenesis. Development of effective treatment therapies is also likely to depend upon biofilm architecture. There is little information, however, concerning the structure of biofilms existing on human tissues in vivo, largely because it is extremely difficult to generate and then harvest such biofilms intact. In an effort to overcome these difficulties, we have developed an intra-

KEY WORDS: biofilm, plaque, architecture, confocal microscopy.

Received July 15, 1998; Last Revision April 20, 1999; Accepted May 7, 1999


21

22

Wood et a/.

J Dent Res 79(1) 2000

oral device (Fig. 1) for generating plaque biofilms in vivo on natural surfaces of human enamel. The device can be removed intact from the mouth for ex vivo study with no disturbance to biofilm architecture. We have already shown that biofilms formed in this way contain populations of bacteria consistent with the range of organisms known to be present in natural dental plaque, in addition to demonstrating the potential for sitespecific chemical and microbial studies (Robinson et al., 1997). Here we report the use of this device for the study of the threedimensional structure of such natural plaque biofilms. Traditionally, electron microscopy has been the method of choice for studying biofilm composition and structure because of its high resolution (Listgarten, 1976; Theilade et al., 1976). However, this approach, because of the necessary dehydration, fixing, embedding, and staining of specimens, may cause considerable distortion of biofilm structure and may change the relationship of one component with respect to another. The advent of confocal scanning laser microscopy (CSLM), while of lower resolution than electron microscopy, has eliminated or considerably reduced these problems. With CSLM, biofilms can be studied in their natural hydrated state, with no requirement for dehydration, fixation, or staining. In addition, the optical sectioning properties of the CSLM mean that very thin optical sections can be taken at increasing depths through the biofilm, free from out-of-focus blurring. Such digitized data can be re-assembled to provide tri-dimensional information. The use of specific fluorescent probes alongside the reflection mode also provides the possibility for the identification of specific matrix components as well as bacteria (Wagner et al., 1994) within the biofilm in addition to the investigation of transport processes (DeBeer et al., 1994a,b; Lawrence et al., 1994; Stoodley et al., 1994). Surprisingly, it is only relatively recently that CSLM has been applied to dental plaque biofilms. In vitro systems have been used for the study of coaggregation in periodontitis (Cook et al., 1998), while the use of enamel and glass slabs in vivo has yielded important information on the viability of microorganisms in plaque biofilms where only bacteria in the outer regions of the biofilm aggregates were viable, as shown by the use of fluorescent probes (Netuschil et al., 1998). In this paper, we report the use of confocal microscopy to determine the structure of natural, intact human dental plaque biofilms formned on enamel substrates. We discuss the possible importance of these structures to plaque behavior, management, and therapy and consider the relevance of this model to the study of other in vivo biofilms.

NATURAL ENAMEL SURFACE NYLON RING ADHESIVE ENAMEL PARTICLE

-2 mm '

MATERIALS & METHODS Generation of Plaque Biofilms Plaque biofilms were generated by means of the Leeds in sitl device, comprised of a nylon ring attached to an enamel substrate, as previously described (Robinson et al., 1997). Ethical approval for the study was obtained from the Leeds Healthcare/United Leeds Teaching Hospitals Trust Research Ethics Committee, and all volunteers participated after giving inforned consent to the protocol. Briefly, two suitable free buccal surfaces were chosen on the first or second upper molars of' each of eight healthy volunteers. The enamel surface was lightly etched with 10% maleic acid for 15 sec, rinsed with water, and dried. A drop of adhesive (HEMA BIS-GMA resin Scotchbond, Multipurpose Dental Adhesive System, 3M, Minneapolis, MN, USA) was placed on the etched surface, and the devices were bonded into place by means of Herculite composite resin. Volunteers were instructed to follow their normal oral hygiene procedure but were asked to take care when brushing near the devices to ensure, as far as possible, that the contents of the devices remained undisturbed. Devices were left in place for four days and were then debonded by means of an orthodontic bracket remover so that the rings would be retained intact with the plaque in sitiu. Immediately upon removal, the devices were placed in reduced transport fluid (RTF), pH 8, and imaged by the confocal laser scanning microscope (CLSM) without any further preparation.

Confocal Laser Scanning Microscopy Non-invasive confocal imaging of intact biofilms was performed with the use of a Noran Odyssey CLSM (Noran Instruments UK, Bicester, Oxfordshire, UK) equipped with an argon ion laser and mounted on a Nikon Optiphot microscope (Nikon UK Ltd., Kingston Upon Thames, Surrey, UK). The objective used was a water immersion "dipping" lens (63x, Zeiss, Carl Zeiss Ltd., Welwyn Garden City, Herts, UK) with a working distance of 1.45 mm. Biofilm architecture was demonstrated with the CLSM used in reflected-light mode. Depth measurements were taken at regular intervals across the width of the devices. This was achieved by manual focusing through the plaque until the enamel surface could be seen. To determine the structure of the biofilms, we took a series of horizontal (xy) optical sections with a thickness of 1.5 .Lm, at 0.5-pm intervals, throughout the full depth of the biofilms. Images were either displayed individually or reconstructed into 3D image datacubes by means of software developed by Mr. C. Eberhardt, Department of Physics and Astronomy, University of Leeds. This software is similar to many other commercially available 3D visualization packages (White, 1995). In addition, we took vertical (xz) sections to determine biofilm thickness and architecture. Negative staining of the biofilms was achieved when the devices were incubated in fluorescein (Sigma Chemiiical Corp., Poole, Dorset, UK) (2 mg/mL) for 15 min followed by fluorescence and reflectance imaging. The excitation wavelength for the fluorescence imaging was 488 nm, and emitted light above 500 nm was collected.

RESULTS -6mm Figure 1. In situ plaque-generating device (adapted from Robinson et a/., 1997).

Upon removal from the volunteers, all 16 of the in situi devices showed evidence of plaque formlation, although the amount of plaque formed varied between individuals. Plaque build-up was


J Dent Res 79(l) 2000

Confocal Microscopy of Plaque Biofilms

23

Figure 2. Reflected-light xy sections of (a) the surface of the enamel underlying the plaque and (b) the plaque 10 p.m above this area. Light reflected from the enamel surface was clearly attenuated by the biofilm, but the enamel surface could still be clearly distinguished by its characteristic pattern of rod end pits larrowed). Scale bar = 10 pLm. heavier at the enamel/ring junction (plaque depth ranged from 75 to 220 p.m) and lighter toward the center of the device (range = 35 to 215 pm). Reflected-light-mode imaging revealed that the plaque was highly heterogeneous in structure. Horizontal (xy) imaging at increasing depths through four-dayold biofilms showed that the enamel, while not providing a highly reflective surface due to the presence of the biofilm, Figure 3. Serial xy optical sections of 1 .5- Lm thickness of intact four-day could still be clearly distinguished by its characteristic pattern human plaque. Sections were taken in reflected-light mode at 15-Vpm small, and large both of rod end pits (Fig. 2). Above this, intervals from the surface of the plaque to a depth of 75 p.m. Bacteria are highly reflective, apparently cellular aggregates were visible, seen in shades of grey to white; dark areas are putative voids and channels. Scale bar = 10 pm. separated by areas of negligible reflectance which were presumed to be fluid-filled channels (Fig. 3). Negative staining of the same biofilms (Fig. 4) confirmed that these were fluidthe voids frequently resulted in the formation of mushroomfilled channels, since they exhibited fluorescence (green, shaped structures which had relatively narrow attachment fluorescence mode) associated with the presence of fluorescein, to the enamel, with the majority of the biomass being in points did mode) reflected-light (red, whereas the bacterial aggregates the upper areas of the biofilm (Fig. 5). not take up the dye. 3D reconstructions of series of xy images (Fig. 6), as well The use of vertical (xz) sectioning (Fig. 5) showed that as demonstrating the open, extended structure of the plaque, some of the channels penetrated the entire depth of the plaque, also illustrated the differences in depth/density of the biofilm giving rise to columns of bacterial/matrix aggregates extending were apparent within the same area of ilterest. which and distribution of The size upward from the enamel surface. Downloaded from jdr.sagepub.com by guest on July 13, 2011 For personal use only. No other uses without permission.


J Dent Res 79(l) 2000


25

Figure 5. Typical vertical (xz) section through a four-day human plaque sample taken in reflected-light mode. Images were taken at 0.6-vLm intervals from the top of the biofilm to the enamel surface underlying it. The image clearly demonstrates the bacterial aggregates (grey-white) separated by areas of low reflectance (arrowed) presumed to be channels. Inverted biomass (M) and associated narrow attac-hment points (A) can also be observed. Scale bar = 25 lam.

processes within the biofilm. Bulk flow through biofilm ACKNOWLEDGMENTS channels has been demonstrated in a three-species in vitro We thank Mr. C. Eberhardt, Departmiienlt of' Physics and biofilm by means of fluorescently labeled latex beads (Stoodley University of Leeds, for his help and advice on the Astronomy, et al., 1994). Similar studies (DeBeer et al., 1994a) used confocal microscope. This work was supported by funding fluorescein microinjection and particle tracking to show that from The Special Trustees of The General liilfrmary at Leeds while liquid flowed fairly freely through channels, it was ST99/08). (Ref: stagnant in cell aggregates. This implies that diffusion was a major factor contributing to mass transfer within cell/matrix REFERENCES clusters, a process which is much slower than the bulk flow in Bowden GH (1990). Microbiology of root surtace caries in huliLanis. J substrate generate could the channels. Such a differential Dent Res 69:1205-1210. concentration gradients from the channels to the interior of the Brown MR, Allison DG, Gilbert P (1988). Resistance of bacterial cell/matrix clusters. This has been supported by further work biofilms to antibiotics: a growthl-rate related effect'? .1 Aotin,icr0ob (DeBeer et al., 1994b) which demonstrated that oxygen tension Chemother 22:777-780. was highest in biofilm channels, whereas the environment in Caldwell DE, Lawrence JR (1988). Study of attached cells in the interior of the clusters was essentially anaerobic. continuous flow slide culture. In: CRC handbook of laboratory Biofilm channels, therefore, may form a rudimentary systems for microbial ecosystem research. Wihmpenniy JWT, model homeostasis circulatory system which would aid in bacterial Boca Raton, FL: CRC Press, pp. 117- 138. editor. channels of "circulatory" The presence (Costerton et al., 1995). Caldwell DE, Korber DR, Lawrence JR (1992). Imaginig ol bacterial in dental plaque reported here will also have important cells by fluorescence exclusion using scanning confocal laser implications for the movement of tooth-damaging organic microscopy. J Microhiol Meth 15:249-261. acids, bacterial toxins, and other antigens as well as for the Cook GS, Costerton JW, Lamont RJ (1998). Biofilimi formation by delivery of therapeutics to the desired targets within the Porphyromonas gingivalis and Streptococcus gordlnnii. J Periodont biofilm. Res 33:323-327. The similarities in structural organization between these in JW, Cheng KJ, Geesey GG, Ladd TI, Nickel JC, Dasgupta systems vitro Costerton in for reported vivo biofilms and many of those (1987). Bacterial biofilms in nature and disease. Ann Rev M, et al. the to common aspects generic are that there support the view Microbiol 41:435-464. generation of all bacterial biofilm structures. Therefore, it is Costerton JW, Lewandowski Z, Caldwell DE. Korber DR, Lappin-Scott possible that our novel in vivo system may also be used to HM (1995). Microbial biofilms. Ann Rev Microbijl 49:711 -745. model transport phenomena in other natural biofilms of Davies DG, Chakrabarty AM, Geesey GG (1993). Exopolysaccharide medical significance where it is normally difficult, if not production in biofilms: substratum activation of alginate gene impossible, to generate intact biofilms for ex vivo study. This expression by Pseldonionas aeruginosa. Appl Environ Microbiol new of development in the would be of tremendous importance 59:1181-1186. of determination in the and treatment biofilm for therapeutics Downloaded from jdr.sagepub.com by guest on July 13, 2011 For personal use D, only. Stoodley No other uses without Lewandowski Z (1994a). Liquid flow in P, permission. DeBeer their efficacy.

:I


J Dent Res 79(1) 2000


Nyvad B, Fejerskov 0 (1989). Structure of dental plaque and the plaque-enamel interface in human experimental caries. Caries Res 23:151-158. Peters A, Wimpenny JWT (1988). A constant depth laboratory model film fermenter. In: CRC handbook of laboratory model systems for microbial ecosystems. Wimpenny JWT, editor. Boca Raton, FL: CRC Press Inc., pp. 175-195. Robinson C, Kirkham J, Percival R, Shore RC, Bonass WA, Brookes SJ, et al. (1997). A method for the quantitative site-specific study of the biochemistry within dental plaque biofilms formed in vivo. Caries Res 3 1:194-200. Singleton S, Treloar R, Warren P, Watson GK, Hodgson R, Allison C (1997). Methods for microscopic characterization of oral biofilms: analysis of colonization, microstructure, and molecular transport phenomena. Adv Dent Res 11:133-149. Sissons CH (1997). Artificial dental plaque biofilm model systems. Adv Dent Res 11:110-126. Stewart PS, Peyton BM, Drury WJ, Murga R (1993). Quantitative observations of heterogeneities in Pseudomonas aeruginosa biofilms. App/ Environ Microbiol 59:327-329. Stickler DJ, King JB, Nettleton J, Winters C (1993). The structure of

27

urinary catheter encrusting biofilms. Cells Mater 3:315-320. Stoodley P, DeBeer D, Lewandowski Z (1994). Liquid flow in biofilm systems. Appl Env Microbiol 60:2711-2716. Theilade J, Fejerskov 0, Horsted M (1976). A transmission electron microscopic study of 7-day old bacterial plaque in human tooth fissures. Arch Oral Biol 21:587-598. van der Hoeven JS, de Jong MH, Rogers AH (1985). Effect of utilisation of substrates on the composition of dental plaque. FEMS Microbiol Ecol 31:129-133. Wagner M, Assmus B, Hartmann A, Hutzler P, Amann R (1994). In situ analysis of microbial consortia in activated sludge using fluorescently labelled, rRNA-targeted oligonucleotide probes and confocal scanning laser microscopy. J Microsc 176:181-187. Walker JT, Mackerness CW, Rogers J, Keevil CW (1995). Heterogeneous mosaic a haven for waterborne pathogens. In: Microbial biofilms. Lappin-Scott HM, Costerton JW, editors. Cambridge, UK: Cambridge University Press, pp. 196-204. White NS (1995). Visualisation systems for multidimensional CLSM images. In: Handbook of biological confocal microscopy. Pawley JB, editor. New York; Plenum Press, pp. 211-254.
