OOa3-9969190 83.00+ 0.00
Archs oral Bid Vol. 35,No. 4,pp. 295-299,1990 Printed in Great Britain. All rights reserved
Copyright c 1990Pergamon Press plc
TH[E DISTRIBUTION AND QUANTITATION OF CIEMENTUM-BOUND LIPOPOLYSACCHARIDE ON PERIODONTALLY DISEASED ROOT SURFACES OF HUMAN TEETH F. J. HUGHES and F. C. SMALES Department
01‘Oral Medicine & Periodontology, The London Hospital Medical College, Turner St, London El 2AD, England (Received 21 April 1989; accepted 31 September 1989)
aim of this study was to estimate the concentration of cementum-bound LPS on a group of 12 teeth that had been extracted because of periodontitis. LPS on scaled root surfaces was labelled by immunogold/silver staining. The concentrations of LPS were estimated by quantifying the amount of bound silver la.bel, using X-ray microanalysis in areas free of plaque or calculus. These were compared against standards of known LPS concentration, which were separately prepared for each sample. Cementum-bound LPS was detected at concentrations of up to 2 EU/mm2 of affected root surface. However, most of the root surfaces had considerably lower concentrations than this, the mean of all samples never exceeding 0.7 EU/mm2. LPS concentrations were highest on cementum towards the apical regions of the affected pocket. These findings confirm that cementum-bound LPS is only present in low concentrations on affected teeth, and suggest that the clinical significance of cementum-associated LPS may have been over-estimated in the past. The demonstration of LPS appears to be more important as an indicator of retained bacteria and calculus than of cementum-bound LPS per se. Key words: cementum, lipopolysaccharide,
It has been suggested that the extensive removal of cementum by root planing is required to eliminate cementum-bound LPS during periodontal therapy. In order to substantiate or refute this claim it is necessary to know the distribution and likely concentrations of both the total cementum-associated LPS and the cementum-bound LPS on teeth affected by destructive periodontal disease. The distinction between total cementum-associated and cementumbound LPS may be significant when evaluating the importance of root planing. In an earlier scanning electron-microscopic study (Hughes et al., 1988), an immunogold/silver staining technique was used to label cementum-associated LPS in situ on extracted tooth samples. Our aim now was to analyse similar tooth samples by X-ray microanalysis to determine the amount of specific gold/ silver-labelled antibody binding to LPS in order to estimate the concentration of cementum-bound LPS and to determine its distribution on affected root surfaces.
INTRODUCTION Lipopolysaccharides are macromolecules that are released from the cell walls of Gram-negative bacteria during active cell growth or when bacteriolysis occurs, and form the prime active components of bacterial endotoxins (Cruchley, Marsh and Cameron, 1967). The presence of LPS associated with root surfaces exposed to a periodontal pocket has been reported in a number of studies (e.g. Aleo et al., 1974; Aleo, De Renzis and Farber, 1975; Jones and O’Leary, 1978; Fine (etal., 1980). This contamination has been referred to as ‘cementum-associated LPS’, as it is not clear whether it is associated with retained bacteria, is adsorbed on to the surface of the cementurn, or absorbed int’a the cementum (Daly, Seymour and Kieser, 1980). A number of studies have suggested that LPS does not penetrate deeply into sub-surface cementum (Nakib et al., 1982; Moore, Wilson and Kieser, 1986; Hughes and Smales, 1986; Maidwell-Smith, Wilson and Kieser, 1987; Cheetham, Wilson and Kieser, 1988). In addition, most of the remaining cementumassociated LPS detectable on affected root surfaces after scaling is due lo LPS in retained bacteria and calculus rather than adsorbed ‘cementum-bound LPS’ (Hughes, Auger and Smales, 1988).
MATERIALS AND METHODS Twelve single-rooted teeth extracted because of severe adult periodontitis were investigated. Prior to extraction all of the teeth showed attachment loss and probing depth measurements of at least 7mm. The 12 patients were aged between 44 and 61 yr (mean Sl), were medically healthy, and had no history of periodontal treatment in the previous 12
EU/mm’, endotoxin units/area of root surface; LAL, Limdus amoebocyte lysate assay; LPS, lipopolysaccharide. 295
J. HUGHES and
months. A single 2mm thick longitudinal slice was prepared from each specimen and scaled to remove all visible plaque and calculus, care being taken to ensure that cementum was not deliberately removed. During this procedure the samples were held with pyrogen-free forceps apical to the level of the remaining periodontal attachment. These slices were subjected to immunochemical analysis to determine the amount of cementum-bound LPS. The remainder of each tooth sample was left unscaled, and was used for the determination of total cementum-associated LPS and for preparation of LPS standards. The pathologically altered surface area of each sample was first estimated by measuring the coronal circumference at the cementum-enamel junction (CC), the apical circumference at the attachment level (AC) and the length between them (L) in mm. The surface area was then calculated by the formula: affected
area = [CC + AC]/2 x L mm’.
The teeth were then thoroughly root planed, and all of the tissue that was removed was carefully collected in pyrogen-free water. LPS was extracted from each of these samples by treatment with 45% aqueous phenol at 68°C for 1 h, and purified by ultracentrifugation at 40,OOOg for 3 h, digestion in 0.1 mg/ml ribonuclease and deoxyribonuclease (Sigma Ltd, Poole, England) at 37°C for 2 h, followed by further ultracentrifugation at 40,OOOg for 3 h. The supernatant was removed and the sample was resuspended in 1 ml of 0.1% bovine serum albumin/carbonate buffer, pH 9.6, and triplicate 100 ~1 portions were assayed for LPS content using a chromogenic LAL (QCL-1000, Whittaker MA Bioproducts, Laboratory Impex Ltd, Twickenham, England). Previous studies using polymyxin B affinity chromatography have shown that a positive LAL reaction with such a preparation is due to the presence of LPS rather than other LAL-positive substances (Moore et al., 1986; Hughes et al., 1986). The mean concentration of total cementurn-associated LPS for each tooth, expressed as EU/mm’, was calculated by dividing by the affected surface area. L PS standards Triplicate 100~1 samples of the LPS preparation described above were coated on to microtitre plate wells without further dilution, and at dilutions of 1: 2 and 1:4; control wells were exposed to coating buffer only. Preliminary experiments, in which the concentration of LPS in the coating buffer had been determined by LAL assay before and after coating, suggested that this regimen resulted in the binding of over 95% of LPS to the wells (data not shown). Each test was done in triplicate. The surface area of the coated wells was estimated to determine LPS concentrations In this way a series of standards of known LPS concentration were separately prepared for each tooth slice.
pooled human serum with a mixed oral LPS sample, obtained from the plaque and calculus of 28 extracted teeth, which had been purified as above and affinity purified using a polymyxin B column. This system was designed to provide antibodies to as wide a range of oral LPS types as possible. Tooth samples were incubated with the human anti-LPS serum, followed by a rabbit anti-human immunoglobulin serum (Dako Ltd, High Wycombe, England) and a colloidal gold-labelled goat anti-rabbit immunoglobulin serum (Janssen Life Sciences Ltd, Wantage, England). Bound gold label was subsequently enhanced with a silver development stage. Samples were then dehydrated and sputter-coated with platinum. The LPS standards of known concentration were processed in the same way and in parallel with their respective tooth sample to prevent any variation in intensity of staining and thickness of sputter-coating between test and standard. Analysis of specimens Teeth and standards were all viewed in a Philips EM 400 unit with a secondary electron detector. X-ray microanalysis was with an EDAX 9100/60 system in conjunction with the Philips EM unit. Silver concentrations on the specimens were determined measuring silver Lcr peaks, using the semi-quantitative program of the EDAX operating system software. All analyses were carried out at x 800 magnification, using a beam current of 18 mA, accelerating voltage of 20 kV, a local tilt angle of 27’ and a take-off angle of 47”. Each sample was analysed for 8 scans of the field, at a scan time of 32 ms/line ( = 32 s/scan). Four different fields were examined on each standard. Tooth samples were examined by moving the tooth apico-coronally from the level of attachment, and EDAX analysis was carried out on every field in that line that consisted exclusively of surface cementum and cuticle. In order that only cementum-bound LPS should be estimated, fields which contained plaque or calculus were not analysed. Each plaque-free field was measured 3 times. The data were expressed as the silver concentrations found on the specimens in counts per second. The results from the standards were used to construct a standard curve for each specimen, plotting detected silver concentration against LPS concentration. Silver concentrations of the test samples (measured as counts per second) were converted to LPS concentrations by linear regression from the corresponding standard curve. The results were expressed as EU/mm’ and represented the amount of cementum-bound LPS. RESULTS
The appearances of the tooth samples when viewed by scanning electron microscope have been described by Hughes et al. (1988) and will not be considered further here.
Concentration of total cementurn-associated LPS
The tooth slice preparations were processed and reacted with the immunogold/silver staining system as described by Hughes et al. (1988). In brief, a human anti-LPS serum was prepared by absorbing
The results of the LAL assay on the LPS samples obtained from the individual teeth showed considerable variation between samples, ranging from 100 to 1108 EU (mean = 527 EU). The mean concentration
Table 1. Mean total cementum-associated LPS and mean and maximum cementum-bound LPS concentrations detected on samples
Cementumassociated LPS (EU/mm2)
Mean cementumbound LPS (EU/mm2)
Max cementumbound LPS (EU/mm2)
Cementumbound LPS of cementumassociated LPS
0.64 3.04 4.87
0.07 0.40 0.55 0.19 0.12 0.26 0.47 0.38 0.61 0.56 0.09 0.35
0.14 0.63 1.8 0.45 0.46 1.42 1.93 0.78 1.24 1.73 0.57 1.09
11 13 II 13 I 14 8 9 14 11 5 I
A B C
D E F G H
1.71 1.92 5.68 4.09 4.41 5.00 1.83 4.91
J K L
of total cementum-associated shown
LPS on each tooth
LPS standards A typical example of a standard curve is shown in Fig. 1. A close linear relationship was consistently seen between silver concentration and LPS concentration. The 3 measurements of silver concentrations made for each field only showed variations of up to 3 counts/s, indicating the high reproducibility of the measurement technique. Non-specific binding to the plastic wells used for the preparation of standards was between 7 and 10 counts/s for all of the samples tested. Because of the need to use the teeth for the preparation of standards, it was not possible to determine directly non-specific binding to the teeth using appropriate immunohistochemical controls. However, in earlier studies using similar specimens and healthy tooth specimens, non-specific binding to the root surfaces was not detected (Hughes et al., 1988). Concentration of cementum-bound LPS Between 18 and 24 different fields were examined on each test sample and the maximum and mean cementurn-bound LPS concentrations were recorded for each tooth sample (Table 1). The maximum concentration from each different sample ranged from 0.14 to 1.93 EU/mm*, and the mean concentrations ranged from 0.07 to 0.61 EU/mm’. The
_ 200 In ‘, Y
8 I, a
Mean k SD
( EU /mm
I 1.0 2
Fig. 1. Typical standard curve obtained by plotting detected silver concentration measured as counts per second against LPS concentration (EU/mm2).
percentage of total cementurn-associated LPS (determined by LAL assay) present as cementurn-bound LPS was also calculated (Table 1). In general, highest concentrations were found towards the more apical aspects of the pathologically altered root surface. With the method described, in which plaque and calculus-free fields were systematically identified on the specimens by moving along the specimen in an apico-coronal direction, it was possible to build up rough profiles of the LPS concentration through an apico-coronal slice of the specimen [Fig. 2(a)]. A mean profile of LPS concentration through an apico-coronal slice from all the samples was determined by taking the LPS concentration for each field, expressed as a percentage of the maximum LPS concentration for that sample [Fig. 2(b)]. Although there was a wide variation between samples, this confirmed the subjective observation of a trend towards relatively higher LPS concentrations at the more apical aspects of the affected surfaces. DISCUSSION
Our findings provide new information about the concentration of LPS on pathologically altered cementum through use of a novel method that allows concurrent scanning electron microscopic examination of the area to be investigated. The advantage of this technique is it distinguishes between LPS associated with bacteria and calculus and that which is cementurn-bound, adsorbed to the pathologically altered root surface. The measurement of cementurn-bound LPS may be important in attempts to evaluate the need for root planing in periodontal treatment. In addition, we could measure intra-surface regional variations of LPS concentration, which allows a maximum local concentration of LPS to be estimated, as opposed to measuring a mean concentration on the affected root surface. Our estimations show that, on average, around 10% of the total LPS remaining on visibly clean root surfaces was cementurn-bound. However, this is only a rough estimate because of possible inaccuracies in the method for measurement of the surface area of
F. J. HUGHESand F. C. SMALES
the sample, and because the calculation assumed that the LPS concentrations of the test slice was also to be found on the remainder of the root surface. We could not estimate the amount of total LPS associated with bacteria and calculus by analysis of silver concentration as the surface irregularities at these sites make accurate quantitative X-ray analysis impossible with this method. Although our results suggest that our techniques are reproducible, the method of assessing LPS concentration is somewhat indirect, and there are a number of potential sources of error, which would suggest that the data should be interpreted as a general guide to the order of magnitude of LPS concentration, rather than ascribing great significance to the actual figures obtained. We cannot be certain that our anti-LPS serum would react with all of the LPS present on any given sample because of possible variations in the bacterial sources of those LPS types present, although our system was designed to incorporate antibodies to as many different types as possible. However, by using standards prepared from the same tooth sample, this problem was minimized, as the ratio of undetectable LPS to that detectable by the antiserum should be the same in both the sample and its autologous standards. The method of EDAX analysis was standardized, but the differences between the physical properties of the coated plastic standards and the tooth specimens may still have had some influence on the silver X-ray emissions produced. On the other hand, as the silver
Coronal to ottochment
Mean f SD
was bound to the surface of the specimens, this effect is not likely to have been very great. In addition, the range of concentrations of LPS standards did not always coincide with some of the LPS concentrations found on the root surfaces, so we had to assume that the standard curve remained linear beyond the range of the standards. Our findings suggest that cementum-bound LPS occurs at maximum concentrations in the order of 1-2 EU/mm’ of affected root surface, with such values being particularly common towards the more apical regions of a periodontal pocket. Much of the root surface of our samples had considerably less LPS than this. There were significant variations between different samples, although there was no apparent correlation between this figure and clinical parameters such as age and sex of the patient, pocket depth, or area of affected root surface (data not shown). The increased concentrations of LPS towards the more apical aspects of the affected cementum might reflect an increase in numbers of Gram-negative bacteria in the more apical regions of periodontal pockets. This finding is consistent with those of Fine et al. (1978) who demonstrated higher LPS concentrations by LAL assay in subgingival than similar supragingival plaque. The total cementum-associated LPS on each root surface after scaling ranged from 100 to 1108 EU (mean = 527 EU). These figures are approximately equivalent to 9-100 ng of Escherichia coli LPS (mean 48 ng): this way of expressing LPS concentration has been used by others working with the LAL assay and our results for the amount of LPS on affected root surfaces are similar to those obtained before. Thus Moore et al. (1986) found that LPS bound to
root surfaces had a mean concentration of 81 ng/ tooth, after rinsing and cleaning with a rotary brush; Maidwell-Smith et al. (1987) found LPS concentrations ranging from 19 to 394ng on individual teeth after similar rinsing and brushing; McCoy et al. (1987) found an average of 89.5 EU on each tooth surface (as opposed to whole root) prior to root planing. The concentrations of cementum-bound LPS that we detected are extremely low when compared with the LPS concentrations used in many of the studies that have sought to ascertain the effects of LPS on fibroblast growth and function in vitro; these have generally found that LPS is cytotoxic to cultured fibroblasts at concentrations in the order of lOOpg/ml (Aleo et al., 1974; Olson, Adams and
Coronal to attachment
3 4 level (mm)
Fig. 2. (a) Typical example of LPS concentrations measured from consecutive fields of approx. 0.25 mm in width from one specimen. Highest LPS concentrations are seen between 0.5 and 2 mm coronal to the attachment level. Spaces in the histogram indicate a field where no reading was taken, due to the presence of plaque and calculus, or alternatively where surface cementum had been removed. (b) Mean and standard deviation of LPS concentrations from all samples from consecutive fields apico-coronally. Data have been normalized by expressing LPS concentrations as the percentage of the maximum LPS concentration detected for each sample.
Layman, 1985; Layman and Dietrich, 1987). Further experiments are required to ascertain if LPS at low concentrations may influence cell attachment and viability. However, our quantitative studies support the view that the clinical significance of cementum-bound LPS has been over-emphasised in the past, and that extensive root planing in order to remove cementum-bound LPS may not be justified. REFERENCES Aleo J. J., De Renzis F. A., Farber P. A. and Varboncoeur A. P. (1974) The presence and biologic activity of cementurn-bound endotoxin. J. Periodont. 45, 672475.
Aleo J. J., De Renzis F. A. and Farber P. A. (1975) In vitro attachment of human gingival fibroblasts to root surfaces. J. Periodonr. 46, 639-645. Cheetham W. A., Wilson M. and Kieser J. B. (1988) Root surface debridemen-an in vifro assessment. J. c/in. Periodont. 15, 288-292. Cruchley M. J., Marsh D. G. and Cameron J. (1967) Free endotoxin. Nature 214, 1052. Daly C. G., Seymour G. J. and Kieser J. B. (1980) Bacterial endotoxin: a role in periodontal disease? J. Oral Parh. 9, I-15. Fine D. H., Tabak L.. Salkind A. and Oshrain H. (1978) Studies in plaque pathogenicity. II. A technique for the specific detection of endotoxin in plaque samples using the Limulus Lysate Assay. J. periodonr. Res. 13, 127-133. Fine D. H., Morris M. L., Tabak L. and Cole J. D. (1980) Preliminary characterization of material eluted from the roots of periodontally involved teeth. J. periodon!. Res. 15, 10-19. Hughes F. J. and Smales F. C. (1986) Immunohistochemical investigation of the presence and distribution of cementum-associated lipopolysaccharides in periodontal disease. J. periodoni’. Res. 21, 660467. Hughes F. J., Auger D. W. and Smales F. C. (1988) Investigation of the distribution of cementum-associated lipopolysaccharides in periodontal disease by scanning electron microscope immunohistochemistry. J. periodont. Res. 23. 100-106.
Jones W. A. and O’Leary T. J. (1978) The effectiveness of in uivo root planing in removing bacterial endotoxin from the roots of periodontally involved teeth. J. Periodont. 49, 337-342. Layman D. L. and Dietrich D. L. (1987) Growth inhibitory effects of endotoxins from Bacteroides gingivalis and infermedius on human gingival fibroblasts in v&o. J. Periodont. 58, 387-392. Maidwell-Smith M., Wilson M. and Kieser J. B. (1987) Lipopolysaccharide (endotoxin) from individual periodontally involved teeth. J. clin. Periodonr. 14, 453456. McCoy S. A., Creamer H. R., Kawanami M. and Adams D. F. (1987) The concentration of lipopolysaccharide on individual root surfaces at varying times following in vivo root planing. J. Periodont. 58, 393-399. Moore J., Wilson M. and Kieser J. B. (1986) The distribution of bacterial hpopolysaccharide (endotoxin) in relation to periodontally involved root surfaces. J. clin. Periodont. 13, 748-75 I. Nakib N. M., Bissanda N. F., Simmelink J. W. and Goldstine S. N. (1982) Endotoxin penetration into root surface cementum of periodontally healthy and diseased teeth. J. Periodont. 53, 368-378. Olson R. H., Adams D. F. and Layman D. L. (1985) Inhibitory effect of periodontally diseased root extracts on the growth of human fibroblasts. J. Periodont. 56, 592-596.