ELSEVIER

www.elsevier.nl/locate/jphotobiol

J. Photochem. Photobiol. B: Biol. SO ( 1999) l-7

Journal of Photo+;mistry

Photobiology B:Biology

An in vitro study of the use of photodynamic therapy for the treatment of natural oral plaque biofilms formed in vivo

Simon Wood ‘,*, Brian Nattress b, Jennifer Kirkham a, Roger Shore ‘, Steven Brookes a, John Griffiths ‘, Colin Robinson a

Received 25 August 1998; accepted IS March 1999

Abstract

Seven-day oral plaque biofilms have been formed on natural enamel surfaces in viva using a previously reported in situ device. The devices are then incubated with a cationic Zn( II) phthalocyanine photosensitizer and irradiated with white light. Confocal scanning laser microscopy (CSLM) of the biofilms shows that the photosensitizer is taken up into the biomass of the biofilm and that significant cell death is caused by photodynamic therapy (PDT). In addition, the treated biotilms are much thinner than the control samples and demonstrate a different structure from the control samples, with little evidence of channels and a less dense biomass. Transmission electron microscopy (TEM) of the in vivo-

formed plaque biofilms reveals considerable damage to bacteria in the biofilm, vacuolation of the cytoplasm and membrane damage being clearly visible after PDT. These results clearly demonstrate the potential value of PDT in the management of oral biofilms. Cl I999 Elsevier Science S..4. All rights reserved.

Kew~~rcl\: Biofilmc; Plaque: Photodynamic therapy

1. Introduction

The oral cavity is heavily colonized by a complex. rela- tively specific and highly interrelated range of micro-organ- isms, including both aerobic and anaerobic species of Gram positive and Gram negative bacteria [ I 1. The teeth, unlike the mucosa, do not desquamate. permitting relatively thick biofilms to build up on the tooth surfaces. Microbial biofilms are made up of bacteria adhering both to each other and/or to surfaces or interfaces and embedded in a matrix of extra- cellular polymeric substances (EPS) [ 21. The presence of pathogenic micro-organisms within plaque can lead to dis- ease states such as caries 131 and periodontal disease 13.51.

Traditional methods for the treatment and/or removal of plaque include mechanical removal such as scraping and the use of antibacterial mouthwashes and dentifrices. The former can sometimes be unacceptable, for example, in the mechano- blistering diseases such as epidermolysis bullosa (EB). cspe- cially where there is no intrinsic tooth dysplasia and poor oral hygiene is a major problem. The use of antibacterials can also

* Corrc\ponding author Tel.: t 44-l 13-233-6 I SC): Fax: +44-l 13-233.

615X; E-mail: S.R.Wood~leeds.r~c,uh

lead to problems of bacterial resistance, a matter ofincreasing concern in medicine [ 61.

A possible alternative to these treatments is photodynamic therapy (PDT) (see Ref. [ 71 for a review). The principle of PDT is that a photoactive dye (photosensitizer) is taken up into cells. When irradiated with light of the appropriate wave- length. the sensitizer is activated and causes cell death through the production of’ active oxygen species. Light and drug alone are non-toxic. so only cells that both contain pho- tosensitizer and receive light are affected by the treatment. Thus there is the opportunity to achieve selectivity and target specific areas of the mouth/plaque with this treatment. In addition, PDT is a non-mechanical therapy and is very simple to apply. Therefore, it would also be of use in treating not only medically compromised individuals such as EB suffer- ers, but also children and, where appropriate, the handi- capped.

The purpose of this study was to determine if PDT does have an effect on the bacterial population and structure of human oral plaque biofilms formed in vivo.

Bacteria in biofilms generally have a different phenotype compared with their planktonic counterparts [8] and grow more slowly 19, IO], a characteristic that may explain their

101 I -13-G/99/$ wc frr,nt matter rr’l IWO Ekvier Science S.A. Ail rtghts reserved PIISlOl I-1314(99)00056-1

2 S. Wood et al. /J. Photorhem. Photohid. B: Bid. 50 (1999) 1-7

increased resistance to antimicrobial agents [ 111. Until now, for the purpose of detailed study, plaque samples have tra- ditionally been harvested by mechanically scraping plaque from the tooth surfaces, inevitably disrupting the biofilm structure. To circumvent this problem we have developed a novel in situ device [ 121, which can be used to harvest intact natural plaque biofilms attached to natural or artificial sur- faces from the oral cavity (Fig. 1) . This permits the study of plaque in its natural state, and offers the possibility of mon- itoring changes in its structure and biochemistry in response to different treatments, in this case, PDT.

The photosensitizer used for this study was a cationic Zn( II) phthalocyanine, the structure of which is shown in Fig. 2. This sensitizer has already been shown to be effective against both Gram positive and Gram negative bacteria in planktonic culture [ 131, but this is the first report of its use against bacteria forming a natural biofilm. Sensitizer uptake and changes in hoth cell viability and the structure of the plaque biofilms resulting from PDT were studied using con- t&al scanning laser microscopy (CSLM) in both reflection and fluorescence mode and transmission electron microscopy (TEM).

2. Materials and methods

2.1. Photosensitizer

Pyridinium Zn( II) phthalocyanine (PPC) was synthe- sized by Mr J. Schofield, Department of Colour Chemistry and Dyeing, Centre for Photobiology and Photodynamic Therapy, University of Leeds. Mass and NMR spectroscopy have shown that the compound is a mixture of components with differing degrees of substitution, the average degree of substitution being two. with the disubstituted derivative as the major component [ 14). The phthalocyanine was dis- solved in water at a concentration of 1 mg/ml and stored at .- 20°C until required.

Devices were constructed according to a previously pub- lished design [ 12 1, Briefly. nylon rings (2 mm id, 500 pm in height) were attached to sniall particles of enamel (6 mm diameter) which had been prepared from the crowns of sound, sterilized extracted teeth. The rings provided a stag- nation site for the formation of plaque biofilms on the natural enamel surface.

Eight volunteers each carried two devices, one to act as a control and one to be treated with PDT. Full ethical approval was given for studies using the in situ device in human vol- unteers. Devices were attached to a free buccal surface of the first or second upper molar teeth of healthy volunteers, The

SURFACE

NYLON RING

ADHESIVE ENAMEL m3nc~

I

-2mm

-6mm

Fig, 1, Diagram of the in situ device used for generating intact biofilms in viva (adapted from Ref. [ 121).

--l+ cl- X=

Fig. 2. Structure of the cationic water-boluble Zn( II) phthalocyanine. The average degree of substitution was two.

tooth surface was lightly etched with 10% maleic acid for 15 s, rinsed with water and dried. A drop of adhesive (HEMA BIS-GMA resin Scotchbond, Multipurpose Dental Adhesive System, 3M) was placed on the etched surface and the in situ device bonded in place using Herculite composite resin. Vol- unteers were instructed to follow their normal oral hygiene procedure, but were asked to take care when brushing near the device to avoid disturbing the plaque within the ring. The devices were left in situ for seven days before being removed carefully with an orthodontic bracket remover and placed in sterile reduced transport fluid, pH 8 (RTF) [ 151.

2.4. Photodynamic therupy

Immediately upon removal from the mouth, devices were incubated with PPC at a concentration of 20 kg/ml in RTF for 1 h. The devices were then washed ( X 3 with RTF), care being taken not to disturb the biofilm, and irradiated at room temperature for 30 min under RTF using white light from a 400 W tungsten filament lamp with an average intensity of 22.5 mW/cm’ in the wavelength range 600-700 nm (corre- sponding to the region of maximal absorption by the phthal- ocyanine [ 161) Control samples were exposed to either light or drug alone. Samples were then either viewed immediately using a confocal microscope or processed for TEM.

2.5.1. Photosensitizer uptuke

Following incubation with a 20 kg/ml solution of PPC in RTF, biofilms were imaged using a BioRad MRC600 CSLM equipped with an argon ion laser. Reflection mode was used to image the bacterial aggregates of the biofilm, while fluo-

rescence-mode imaging using the rhodamine filter block (excitation wavelength = 5 14 nm; emission > 550 nm) was found to be suitable for visualizing photosensitizer-associated fluorescence. Corresponding fluorescence and reflection images were then overlayed to demonstrate PPC localization relative to the biomass of the biofilm.

2.5.2. Cell killing The amount of cell death caused by PDT was indicated by

the use of the fluorescent probe propidium iodide (PI), which

is only taken up into dead cells [ 17,181. Control and treated biofilms were incubated with PI (40 kg/ml in RTF) for 15 min before imaging on the BioRad MRC600 CSLM in fluo- rescence mode using the rhodamine filter block. Images were processed using the LaserSharp software provided with the CSLM.

2.5.3. Biojilm structure Control and PDT-treated intact biofilms were placed in

RTF and viewed using a Noran Odyssey CSLM in reflection mode. The microscope was equipped with a Zeiss axioplan X 63, long working distance. water immersion (dipping) objective. Series of horizontal ( XY) images were taken at 0.6 km intervals throughout the depth of both treated and control biofilms and reconstructed into three-dimensional datacubes using software developed by Mr C. Eberhardt. Department of Physics and Astronomy, University of Leeds. Vertical (x:) sections were also used to determine differences in thickness of the biofilms before and after PDT.

2.6. Trunsmission electron microscopy (TEM)

Control and PDT-treated devices were placed in 3.6% glu- taraldehyde in 0.2 M cacodylate buffer, pH 7.4 containing 1 mg/ml Ruthenium Red for I h at room temperature. After

Fig. 3. CSLM imaging showing uptake of PPC into the biomass of the biofilm. Reflective cellular material was located using reflection-mode imag- ing (A J and PPC fluorescence was detected in fluorescence mode using the rhodamine filter block ( 6). An overlay of the two images (C) shows aclose a\\ociation between the phthalocyanine and the biofilm biomass.

Fig. 4. Fluorescence-mode CSLM images taken i0 pm below the surface of propidium iodide-stained biofilms before (A) and after (B) PDT treatment. More intense and widespread staining following treatment is indicative 01 ccl1 killing caused by PDT. Broken lines indicate the position of the nylon ring on each device. Bar = IO pm.

4 S. Wood et ~1. /J. Photorhetn. Photohid. B: Bid. 50 (1999) 1-7

washing the samples were post-fixed and stained in 1% osmium tetroxide, 1 mg/ml Ruthenium Red in 0.2 M phos- phate buffer for 24 h. The devices were then washed, dehy- drated through graded alcohols and embedded in methacrylate resin. Ultrathin vertical (longitudinal) sections representing the entire thickness of the plaque samples were cut using a diamond knife, picked up on copper grids and viewed using a Philips 400 transmission electron micro- scope.

3. Results and discussion

Previous work has shown that PDT is capable of killing oral bacteria in planktonic culture [ 191, plaque scrapings [ 201 and as artificially formed biofilms [ 211 in vitro. In the present study we have used confocal microscopy and trans- mission electron microscopy to investigate the effect of PDT

on plaque biofilms formed in vivo using an in situ device which allows the harvesting of intact plaque biofilms from human volunteers.

CSLM imaging following incubation with PPC for 1 h showed a close correlation between the reflection and fluo- rescence images (Fig. 3)) demonstrating that PPC was taken up into the biomass of the biofilm. Although it is difficult to conclude from these results precisely which structures had taken up the drug (i.e., cells or EPS), the sensitizer was not readily washed out of the biofilm, indicating a close associ- ation between sensitizer and biofilm. Widespread cell killing of bacteria was indicated by staining with propidium iodide after irradiation with light (Fig. 4). Much more intense stain- ing was seen following irradiation. No apparent selectivity was achieved in the killing of the bacteria as evidenced by the fact that staining of the biofilm was reasonably uniform following treatment. This suggests that both aerobes and anaerobes can be affected by PDT. This finding is particularly

Fig. S Lon cytopl urn.

itudir 1al 1 ‘EM micr ographs of con trol (A +B) and PDT ‘-treated ( ;hich hec,

‘C + D) biofilms showing vacuolation (arrowed) and cc ame more electrol n dl ense follow ing PDT. Magr lification : A+C= 10000: B+D=50000.

an of bacterial

S. Wood et al. /.I. Phowohem. Photohid. B: Bid. 50 (1999) 1-7 5

interesting, as it indicates that PDT with PPC is effectiveeven in the absence of oxygen. This could have significant clinical importance, particularly for the treatment of obligately anaer- obic periodontopathogenic bacteria.

When the irradiated plaque biofilm was examined at higher resolution using TEM (Fig. .5), obvious damage to individual bacterial cells was discernible. Vacuolation and condensation of the cytoplasm was the most evident damage seen, with cytoplasm apparently retracted from the cell wall/mem- branes. The membranes themselves had a different appear-

ante following PDT, appearing to be more defined than the control samples.

CSLM was used to investigate the effect of PDT on biofilm structure throughout the depth of the biofilm. As well as allowing sharp imaging of the full thickness of the sample via optical sectioning, CSLM also abolishes the need for fixation of the sample, which in turn means that, unlike SEM or TEM, the sample is free from preparation artefact. Three- dimensional data cubes were also constructed from a series of 160 optical sections, each of 0.6 pm thickness. Our pre-

0 IOOpm Fig. 6. Three-dimensional CSLM reconstruction of a control intact plaque bioiilm based on a series of horizontal (xy) images taken in reflected-light mode from the surface of the plaque (s) through to the enamel surface t e). ( A) Plaque biofilm viewed from the side; (B) biotilm viewed from above. The Figure demonstrates the heterogeneity of structure of the plaque and the presence of channels (arrowed) separating regions of high-density biomass (*) The colour scale varies from yellow ( material closest to the viewer) to red (material further away). Cube dimensions: IO0 X IO0 X 1 SO Pm.

0 1OOpm Fig. 7. Three-dimensional CSLM reconstruction of a PDT-treated biofilm based on a series of horizontal (.rv) images taken in reflected-light mode. The biofilm i\ approximately half the thickness of the controls and shows no evidence of discrete channels. The structure as a whole seems to be less dense than the controls. with high-density bacterial aggregates being replaced by columns of bacteria (c) extending upwards fromthe enamel surface. Cube dimensions: I00 X IO0 X 1.50

**m.

6 S. Wood et al. /J. Photochem. Photobiol. 8: Biol. 50 (199Y) 1-7

vious studies using CSLM have revealed that plaque biofilms formed over a period of four days exhibit a heterogeneous structure characterized by the presence of voids and channels surrounding bacterial aggregates (Fig. 6). Some of these spaces extended through the entire thickness of the biofilm from the external, salivary surface to the enamel substrate.

Following treatment of the biofilms with PDT in this study, the samples were reduced to approximately half the thickness of controls (Fig. 7). The plaque remaining on the enamel was less dense and seemed to be made up of columns of bacterial aggregates extending upwards from the enamel sur- face. The reasons for these changes are not immediately apparent, but it would appear that bacterial membrane dam- age as indicated by TEM may reduce cell-to-cell or cell-to- matrix binding, causing loss of adhesion within the biofilm and subsequent loss of bulk biofilm. In addition, binding of the cationic photosensitizer to the EPS itself may result in significant damage to the bacterial cells embedded in it.

Loss of plaque substance caused by PDT is clearly bene- ficial and amounts to atraumatic scaling. In addition, disrup- tion of plaque structure will have important consequences for homeostasis within the biofilm as transport into and out of the biolilm will be affected. Clearly PDT using this cationic phthalocyanine can cause reduction of plaque biofilms and may be of value as a non-invasive scaling procedure. How- ever, it may not be desirable to kill the entire oral flora, as this would leave the patient open to opportunistic infections. Rather it may be preferable to target the bacteria that have been identified as playing a major role in dental caries, such as mutuns streptococci [ 31. Such preferential killing may then facilitate recolonization by less pathogenic bacterial spe- cies. Work is currently being carried out to determine the degree of selectivity shown by PPC with aview to developing more selective photosensitizers. In addition, light and dosim- etry studies must be completed in order to optimize treatment times [ 221 and minimize any possible normal tissue damage 1231.

4. Conclusions

In conclusion, we have shown that natural human oral biofilms can be disrupted and the bacteria within them dam- aged by PDT using a cationic phthalocyanine. Subject to light and drug dosimetry studies, it is hoped that PDT can be developed to become an alternative treatment for plaque in situations where other therapies are unsuitable, such as in cases of antibiotic resistance and in the mechano-blistering disorders where mechanical removal of plaque is not toler- ated. The model system used here may also be of use for indicating the efficacy of PDT on other biofilms of medical significance, such as those which form on catheters.

Acknowledgements

John Griffiths thanks Yorkshire Cancer Research for finan- cial support for the development of the photosensitizer used in this study.

References

111

[21

[31

[41

[51

161

[71

181

[91

1101

1111

1121

I131

[I41

[I51

1161

[I71

P.D. Marsh, D.J. Bradshaw, Dental plaque as a biolilm, J. Ind. Micro- biol. 15 (1995) 169-175. J.W. Costerton, Z. Lewandowski, D.E. Caldwell. D.R. Korber, H.M. Lappin-Scott, Microbial biofilms, Ann. Rev. Microbial. 49 ( 1995) 711-745. W.J. Loesche, Role of Streptococcus mutans in human dental decay. Microbial. Rev. 50 ( 1986) 353-380. W.J. Loesche, S.A. Syed, B.E. Laughton. J. Stall, Bacteriology of acute necrotising ulcerative gingivitis, J. Periodontology 53 ( 1982) 223-230. W.E.C. Moore, L.V. Holdernan, E.P. Cam. R.M. Smibert, J.A. Burmeister, R.R. Ranney, Bacteriology of moderate (chronic) peri- odontitis in mature adult humans, Infect. Immun. 42 ( 1983) 5 IO-5 15. P. Courvalin, Evasion of antibiotic action by bacteria, J. Antimicrob. Ther. 37 (1996) 855-869. J.M. Nauta, H.L.L.M. Vanleengoed, W.M. Star, J.L.N. Roodenburg. M.J.H. Witjes, A. Vermey, Photodynamic therapy of oral cancer - a review of basic mechanisms and clinical applications, Eur. J. Oral Sci. 104 (1996) 69-81. B.D. Hoyle, L.J. Williams, J.W. Costerton, Production of mucoid exopolysaccharide during the development of Pseudomonas aerugi- nosa biofilms, Infect, Immun. 61 (1993) 777-780. A.G. Gristina, CD. Hobgood, L.X. Webb, Q.N. Myrvik, Adhesive colonisation of biomaterials and antibiotic resistance, Biomaterials 8 (19X7)423-426. J.C. Nickel, I. Ruseka, J.B. Wright, J.W. Costerton, Tobramycin resis- tance of Pseudomonas aeruginosa cells growing as a biotilm on uri- nary catheter material, Antimicrob. Agents Chemother. 27 (1985) 619-624. M.R.W. Brown, D.G. Allison, P. Gilbert, Resistance of bacterial bio- films to antibiotics: a growth-related effect’!, J. Antimicrob. Chemother. 22 ( 1988) 777-780. C. Robinson, J. Kirkham, R. Percival, R.C. Shore, W.A. Bonass, S.B. Brookes, L. Kusa, H. Nakagaki, K. Kato, B. Nattress, A method for the site-specific study of the biochemistry within dental plaque bio- films formed in vivo, Caries Res. 31 ( 1997) 194-200. A. Minnock, D.1. Vernon, J. Schofield, J. Grifftths, J.A. Parish, S.B. Brown, Photoinactivation of bacteria. Use of a cationic water-soluble zinc phthalocyanine to photoinactivate both Gram-negative and Gram-positive bacteria, J. Photochem. Photobiol. B: Biol. 32 ( 1996) 159-164. J. Griffiths, J. Schofield, M. Wainwright, S.B. Brown, Some obser- vations on the synthesis of polysubstituted zinc phthalocyanine sen- sitisers for photodynamic therapy, Dyes and Pigments 33 ( 1997) 65- 78. C.I. Hoover, E. Newbrun, Survival of bacteria from human dental plaque under various transport conditions, J. Clin. Microbial. 6 (1977)212-218. D.J. Ball, S.R. Wood, D.I. Vernon, J. Griffiths,T.M.A.R. Dubbelman. S.B. Brown, The characterisation of three substituted zinc phthalo- cyanines of differing charge for use in photodynamic therapy. A com- parative study of their aggregation and photosensitising ability in relation to mTHPC and polyhaematoporphyrin. J. Photochem. Pho- tobiol. B: Biol. 45 ( 1998) 28-35. S.C. Williams, Y. Hong, D.C.A. Danavall, M.H. Howard-Jones, D. Gibson, M.E. Frischer. P.G. Verity, Distinguishing between living and non-living bacteria: evaluation of the vital stain propidium iodide

S. Wood et (11. /J. Photochem. Phntobiol. B: Biol. 50 (1999) 1-7 I

and its combined use with molecular probes in aquatic samples. J. Microhiol. Methods 32 ( 1998) 225-236.

1181 K.L. Swope, M.C. Flicklinger. The use of confocal laser scanning microscopy and other tools to characterize Escherichia coli in a high- cell-density synthetic hiofilm, Biotechnol. Bioeng. 52 ( 1996) 340- 356.

[ 191 T. Burns, M. Wilson. G.J. Pearson, Killing of cariogenic bacteria by light from a gallium aluminium arsenide diode laser, J. Dent. 22 ( 1994) 273-278.

[ 201 M. Wilson, T. Burns, J. Pratten, G.J. Pearson, Bacteriainsupragingival plaque samples can he killed by low power laser light in the presence of a photosensitiser. J. Appl. Bactetiol. 78 ( 1995) 569-574.

]21 I M. Wilson, T. Bums, J. Pratten, Lilling of Streptococcus sanguis in hiofilms using a light-activated antimicrobial agent, J. Antimicroh.

Ther. 37 ( 1996) 377-38 I

]22] M. Bhatti, A. MacRobert, S. Meghji, B. Henderson, M. Wilson, Effect of dosimetric and physiological factors on the lethal photosensitisation of Porphyromonas gingivulis in vitro, Photochern. Photohiol. 65 (1997) 1026-1031.

]23] N.S. Soukos, M. Wilson, T. Bums, P.M. Speight, Photodynamic effects of toluidine blue on human oral keratinocytes and fihrohlasts and Streptococcus sunguis evaluated in vitro, Lasers Surg. Med. I8 (1996) 253-259.