Critical Reviews http://cro.sagepub.

com/ in Oral Biology & Medicine

Effect of Whole Saliva Components on Enamel Demineralization In Vitro

J.D.B. Featherstone, J.M. Behrman and J.E. Bell CROBM 1993 4: 357 DOI: 10.1177/10454411930040031401 The online version of this article can be found at: http://cro.sagepub.com/content/4/3/357

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

On behalf of:
International and American Associations for Dental Research

Additional services and information for Critical Reviews in Oral Biology & Medicine can be found at: Email Alerts: http://cro.sagepub.com/cgi/alerts Subscriptions: http://cro.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav

>> Version of Record - Jan 1, 1993 What is This?

Downloaded from cro.sagepub.com at UNIVERSITAS GADJAH MADA on December 24, 2013 For personal use only. No other uses without permission.

Critical Reviews in Oral Biology and Medicine, 4(3/4):357-362 (1993)

Effect of Whole Saliva Components Demineralization In Vitro

J. D. B. Featherstone, J. M. Behrman, and J. E. Bell

on

Enamel

Department of Oral Sciences and Rochester Cariology Center, Eastman Dental Center, 625 Elmwood Avenue, Rochester, NY 14620
ABSTRACT: The aim of the present study was to use an in vitro enamel demineralization model (1) to confirm that whole saliva pretreatment conferred acid resistance to dental enamel and (2) to determine whether this phenomenon was attributable to specific salivary proteins, minerals, lipids, or some combination of these. Crowns of human teeth, each with one exposed window, were prepared in groups of ten. They were each pretreated by immersion individually in 4 ml of either (1) clarified whole saliva for 18, 72, or 168 h, (2) dialyzed saliva (3500 MWCO membrane), (3) the "flow-through" fraction from a DEAE separation of whole saliva (neutral and basic proteins), (4) the "eluted" fraction of a DEAE separation of whole saliva (anionic proteins), or (5) a combination of salivary lipids and the DEAE "flow-through" fraction of whole saliva (neutral and basic proteins). Control groups were group 6 with no pretreatment, group 7 pretreated for 168 h in a borate buffer (5 mmol/l), and group 8 pretreated in a mineral solution containing calcium (0.7 mmol/1) and phosphate (2.6 mmolfI). The crowns were then demineralized for 7 d in vitro (0.1 mol/I acetate, 1 mmol/I Ca and phosphate, pH 5.0) to produce artificial caries-like lesions. Lesions were assessed by cross-sectional microhardness profiles, and mineral loss (AZ, Rim x vol% mineral) calculated. Mineral loss (AZ) values decreased linearly with the square root of time of pretreatment by whole saliva, confirning a time-dependent protective effect of salivary pellicle against demineralization of enamel. Pretreatments (168 h) by whole saliva (group 1), dialyzed saliva (group 2), and lipid/'flow through" proteins (group 5) gave equivalent protection (approximately 55%). However, no protection was provided by DEAE-separated protein fractions (no lipid present) or by the mineral alone. The protection of surface enamel against demineralization appears to be given by a combination of specifically adsorbed salivary lipids and proteins.
KEY WORDS: enamel demineralization, saliva, salivary pellicle.

1. INTRODUCTION
The importance of saliva as a protective factor in tooth decay is well documented. When salivary flow is decreased or lost completely, there is a significant increase in decay of all exposed tooth surfaces (Mandel, 1974, 1979). This is very apparent in xerostomic patients who become prone to rampant decay (Bertram, 1967). Selective adsorption of specific salivary proteins to the enamel surface results in the formation of so-called salivary pellicle (Meckel, 1965; Armstrong, 1967; Eggen and Rolla, 1983: Jensen et al., 1992). The inhibition of demineralization of enamel by salivary "pellicle" in vitro has been demonstrated previously (Zahradnik et al., 1976, 1977, 1978; Zahradnik, 1979), but not quantified. The reasons for the observed inhibition of demineralization have not been established. The primary aim of the present study was to use an in vitro enamel demineralization model (1) to confirm that whole saliva pretreatment conferred acid resistance to dental enamel and (2) to determine whether this

phenomenon was attributable to specific salivary proteins, minerals, lipids, or some combination of these. The secondary aim was to quantify the effect of the artificially formed salivary pellicle on the rate and severity of caries-like demineralization of teeth in
vitro.

II. MATERIALS AND METHODS

A. Tooth Preparation
Crowns of extracted human teeth (molars or premolars) were removed from the roots and sectioned, leaving either buccal or lingual surfaces intact. The teeth were cleaned with warm detergent solution (in deionized water) according to standard practice (Featherstone et al., 1983) and assessed as caries free by inspection under a stereo microscope. The crowns were painted with acid-resistant varnish, leaving a window approximately 3 x 2 mm exposed
on each.

1045-4411/93/$.50 1993 by CRC Press, Inc.

357
Downloaded from cro.sagepub.com at UNIVERSITAS GADJAH MADA on December 24, 2013 For personal use only. No other uses without permission.

B. Pretreatment by Saliva and Salivary Components 1. Whole Saliva Pellicle Formation

In the first phase of this study, our aim was to demonstrate whether pellicle derived from whole saliva would impart resistance to enamel against demineralization by acid under in vitro conditions and whether the observed resistance was a function of the time of formation of the artificial pellicle. Stimulated whole saliva was collected on ice by chewing paraffin (Parafilm, American Cyanamid Corp.) and clarified by centrifugation for 10 min to precipitate bacteria and other sediment. The supematant (with PMSF added as an enzyme inhibitor) was used for these experiments. Sectioned crowns were suspended individually in test tubes each containing 4 ml of clarified whole saliva at 37C. The crowns were left immersed in the whole saliva for total periods of 18 h, 3 d (72 h), or 7 d (168 h) to permit pellicle growth. Fresh saliva was used daily. Ten sectioned crowns were individually subjected to each of these conditions and ten additional crowns were not exposed to saliva, serving as controls with no pellicle growth. Each crown was then subjected to an artificial caries-like challenge by immersion in 40 ml of a demineralizing solution for 7 d as described in the following section.

b. DEAE Separation: "Flow-Through" Fraction

Dialyzed saliva prepared as previously described was passed through a DEAE column to separate the anionic proteins from the basic and neutral proteins. We hypothesized that the anionic proteins would adsorb strongly to the enamel surface and may thereby impart protection against subsequent demineralization. The "flow-through" fraction contained the basic and neutral salivary proteins. We also determined by thin layer chromatography (TLC) (samples were run in chloroform/methanol/acetic acid and detected by exposure to iodine) that the DEAE column treatment removed the lipids from the saliva.

c. DEAE Separation: "Eluted" Fraction

The proteins remaining on the DEAE column from group b. were eluted by sodium chloride. This fraction contained the anionic proteins and no lipids (confimned

by TLC).

d. Salivary Lipids
Salivary lipids (total lipids) were extracted from clarified whole saliva prepared as described earlier using a separating funnel and extracting three times with 3:1 chloroform/methanol. The lipid extract (contained in the chloroform/methanol phase) was evaporated to dryness and reconstituted by prolonged shaking in DEAE/'flow-through" proteins in solution (see group b.) to produce a lipid solution in the same volume of fluid that existed in the original saliva. In this way, we prepared a suspension of total salivary lipids at the same concentration that they existed in the original saliva in a vehicle similar to whole saliva but lacking the minerals and anionic proteins. The presence of the lipids was confirmed by TLC as previously described.
e. Control - No Pretreatment
One group of ten teeth received no pretreatment, only 7 d demineralization, as described later.

2. Pretreatment by Salivary Components

In the second phase of the study we chose the 7-d (168 h) pretreatment period to study the possible inhibitory effects of pretreatment of teeth by salivary components separated as we will describe next. We wished to determine whether the protective effect observed for whole saliva could be attributed to proteins, minerals, lipids, specific groups of proteins, or some combination of salivary components. The pretreatment materials used were as follows:

a. Dialyzed Safiva
Clarified whole saliva prepared as previously described was dialyzed in a 3500 MWCO dialysis bag against a 5 mmol/l borate buffer at pH 7.0. The dialysis essentially removed all inorganic ions, in particular calcium, phosphate, and fluoride (confirmed by analy-

f. Control - Borate Buffer

One group of 10 teeth was pretreated in 5 mmol/l borate buffer (pH 7.0) to determine whether the buffer used for the dialysis had any effect on its own.

sis).

358
Downloaded from cro.sagepub.com at UNIVERSITAS GADJAH MADA on December 24, 2013 For personal use only. No other uses without permission.

g. Mineral Only: Calcium and Phosphate

Clarified whole saliva (as previously described) was analyzed for calcium (atomic absorption) and phosphate (molybdate method). A mineral solution (pH 7.0) was made with the same phosphate concentration (2.6 mmol/l) as the whole saliva and with half of the total calcium concentration (0.7 mmol/l). The calcium was used at this concentration to approximate the free calcium available in saliva (approximately 50% of total).

cross-sectional microradiography. Larger values indicate more mineral loss. Mineral loss is directly proportional to the AZ value.

Ill. RESULTS
Subsurface caries-like lesions, as assessed by crosssectional microhardness profiles (Figure 1) were produced in the chemical demineralization system used in this study in all cases, with or without pretreatment by saliva or salivary components.

h. Pretreatment Procedure
Groups of ten teeth were treated in one of the preparations (groups a. through g.) for a total of 168 h. Each tooth was immersed individually in 4 ml of test solution at 37C. Each solution was refreshed daily. At the end of the pretreatment, each tooth was subjected to a 7-d demineralization challenge as described next.

A. Pellicle Formation vs. Time - Whole Saliva

Lesion cross-sectional profiles (mean for each group) are shown in Figure 1 for the whole-saliva experiments. Each point on each line is the mean of the volume percent values for all teeth in the appropriate test group at that depth from the outer surface. The results illustrate the partial protection afforded against demineralization by the saliva pretreatments. Teeth were most demineralized with no pellicle protection. With increased pretreatment exposure time to saliva supematant, the lesion severity decreased. The mean (standard error) relative mineral loss (AZ) values for the whole saliva vs. pretreatment time experiments are plotted in Figure 2. A good linear relationship between square root of treatment time (time of artificial pellicle formation) and mineral loss was observed with a correlation coefficient, r = -0.984 for the linear regression line (Equation 1).
AZ = -199 (sqrt time) + 4493

C. Demineralization Challenge
Artificial caries-like lesions can be produced in one of numerous chemical systems in the laboratory. We chose to use a partially saturated (with respect to calcium and phosphate, i.e., hydroxyapatite) acid buffer. Artificial caries-like lesions can also be produced in acidified gelatin gels or synthetic polymer gels (e.g., White, 1987). We wanted to ensure that there was no interference by the gel or polymer with the artificial pellicles formed during the previously mentioned pretreatments. Therefore, a simple calcium/phosphate/acid buffer was chosen. We used a solution consisting of 0.1 mol/l acetate (added as acetic acid) incorporating 1 mmol/l calcium and phosphate (as CaHPO4), at pH 5.0. All teeth were immersed individually in 40 ml of the buffer for a period of 7 d at 370C.

(1)

B. Salivary Components and Protection Against Demineralization

The mean (standard deviation) relative mineral loss (AZ) values for each of the 7-d pretreatment groups (whole saliva and groups a. through g.) are given in Table 1. Artificial pellicle formed by either (1) whole saliva, (2) dialyzed saliva, or (3) the DEAE "flowthrough" fraction plus lipid added back each provided significant protection (approximately 55 to 60% reduction in lesion severity) against subsequent demineralization. The differences among these three groups were not statistically significant (p <0.05). None of the other pretreatments by minerals alone, protein fractions alone (neither anionic, nor basic and neutral), or borate buffer provided more than minimal partial

D. Lesion Assessment
After exposure to the demineralizing solution, each crown was sectioned through the lesion and embedded in epoxy resin with the cut surfaces exposed according to our established procedures (Featherstone etal., 1983; White and Featherstone, 1987). The embedded teeth were then serially polished and submitted to detailed
cross-sectional microhardness profiling. Relative mineral loss as AZ (gm x vol % mineral) was calculated from the data (White and Featherstone, 1987) to give values comparable to those that would be found by

359
Downloaded from cro.sagepub.com at UNIVERSITAS GADJAH MADA on December 24, 2013 For personal use only. No other uses without permission.

100

80

4+- DOmin only

so
=

-0-

IIhr pdle

40

-+- 3 day p.lklb

-I-

7 day peuck.

50

100

150

200

250

300

Depth from srflace (Pm)

FIGURE 1. Microhardness profiles as volume % mineral vs. depth from the enamel surface for groups pretreated with clarified whole saliva 18 h (U), 72 h (*), 168 h (V), and no pretreatment (control) (+). Each point represents the mean for each group of teeth for observations at each depth from the surface.

'I
x

3,000

E
N 4

1gow
0

10
Root of
lim

15

Squar

(hr)

FIGURE 2. Square root of pretreatment time vs. mineral loss (AZ, whole saliva pretreatments for 18, 72, and 168 h.

im x vol % mineral) for the clarified

360
Downloaded from cro.sagepub.com at UNIVERSITAS GADJAH MADA on December 24, 2013 For personal use only. No other uses without permission.

TABLE 1 Mineral Loss (AZ, gm x vol % Mineral) for Dental Enamel (a) Pretreated by Saliva or Salivary Components (168 h), Followed by (b) 7-d Demineralization (n = 10 per Group)
Pretreatment Dialyzed saliva DEAE flow-through (basic and neutral proteins) plus lipidb Whole saliva DEAE eluted (acidic proteins) Borate buffer pH 7.0 DEAE flow-through (basic and neutral proteins) Minerals only (Ca, P) Demin only - no pretreatment
a

AZ (SD), gm x vol %
1761 1945 1952 3191 3468 3636 3795 4479

(1374)a]
(1874)

(1186)11
(1962)

(1867)] (1442) (1152) (1314)

Values joined by vertical lines not significantly different (p <0.05) by the LSD test. Lipids added back at the same concentration as in the original saliva.

protection against demineralization in this model. These groups did not differ significantly from one another (p <0.05).

IV. DISCUSSION
A clear inhibitory effect on the rate and severity of enamel demineralization by salivary pellicle derived from whole saliva has been demonstrated in this study. This effect was dependent on the time of development of the pellicle. Increasing the exposure time of the enamel to saliva decreased lesion depth and severity of demineralization. Zahradnik and co-workers (1976, 1977, 1978, 1979) previously demonstrated a similar protective effect of pellicle derived from whole saliva, using qualitative microradiography to compare the lesions. Their studies also indicated a probable time dependence. The results of the first phase of the present study confirm those of Zahradnik and co-workers and also provide quantitative evidence that the inhibition is large and is dependent on time, or the square root of time, of pellicle development. No explanation is available so far as to why the observed protection should be dependent on treatment time. Several studies (e.g., Eggen and Rolla, 1983; Lie, 1975; Jensen et al., 1992) have shown that pellicle forms within 2 h on enamel or hydroxyapatite as assessed by protein adsorption measurements. It is not obvious, therefore, why it takes several days before this adsorbed layer achieves its maximum pro-

tective potential against demineralization of the underlying enamel. This may be a result of one or more of several factors including the possibilities of secondary protein adsorption, rearrangement of lipophilic regions of the protein molecules, development of some specific secondary or tertiary structure, adsorption of other molecules, or the direct effect of calcium and phosphate from saliva on the underlying enamel. Prior to the present study, no information was available to establish which, if any, of these factors play a role. To differentiate between the protective effect imparted by adsorbed organic molecules and a possible protective effect imparted by prolonged exposure to calcium and phosphate, the 7-d pretreatments of enamel by dialyzed saliva or mineral solution were carried out. After free calcium and phosphate were removed from whole saliva by dialysis, the remaining organic molecules (proteins and lipids) imparted the same resistance to demineralization as whole saliva. These results indicate that when adsorbed to dental enamel, specific nondialyzable (3500 MWCO) organic components impart resistance against demineralization. In contrast, the 7-d pretreatment of enamel in a solution containing the same concentrations of calcium and phosphate present in the original whole saliva did not impart resistance against subsequent demineralization (Table 1). This study does not address in any way the importance of the presence of minerals during a demineralization challenge. The importance of saturation with respect to hy-

361
Downloaded from cro.sagepub.com at UNIVERSITAS GADJAH MADA on December 24, 2013 For personal use only. No other uses without permission.

droxyapatite and fluorapatite is well established (see article by ten Cate and Featherstone, 1991). In the present study, however, we examined the role of minerals as regards possible protection against demineralization, if the enamel was pretreated in mineral solution, and found no additional protection by this mechanism. The purpose of the DEAE separations of the proteins was to determine whether either of the two major protein groups prepared in this way provided similar protection to that provided by whole saliva. It is well established that the anionic proteins adsorb to apatite and enamel surfaces and inhibit crystal nucleation (Hay and Moreno, 1989). On this basis, we expected that the eluted fraction (containing the anionic proteins) might form a pellicle that provided at least partial protection against subsequent demineralization. However, any proteins that were adsorbed did not provide measurable protection in this model. Further, the "flowthrough" fraction with the neutral and basic proteins provided no measurable protection compared with the borate buffer used as the control. Although we did not examine individual purified protein molecules, these results indicate that there is no one specific protein or group of proteins that, on their own, account for the results observed with whole and dialyzed saliva. It is possible that some combination of proteins separated by us in these experiments could provide a protective effect, but this remains to be investigated. The removal of the lipids by the DEAE column appears to be important. When the total lipid extract was reconstituted into the nonactive "flow-through" DEAE protein fraction at the same lipid concentration present in the original saliva, the protection observed for whole saliva was restored. Further studies will be necessary to determine whether lipids alone (suspended, perhaps, as vesicles in a suitable vehicle) or some specific lipid in conjunction with proteins will have a similar effect. It was beyond the scope of the present study to investigate the lipid effect further. However, the results highlight the potential importance of salivary lipids as part of the caries protective mechanism of the human body. In conclusion, this study highlights the importance of salivary pellicle as part of our cariespreventive processes. The important role of lipids and the combined effect of salivary organic components has been highlighted by these studies.

ACKNOWLEDGMENTS
The authors gratefully acknowledge the assistance of Ms. Carol P. Shields, Dr. Lyndon F. Cooper, and Dr. Mansour Shariati in this work. This study was supported by NIH/NIDR grant DE07003, Rochester Cariology Center.

REFERENCES
Annstrong, W. G.: The Composition of Organic Films Formed on Human Teeth. Caries Res. 1:89-103 (1967). Bertram, V.: Xerostomia: Clinical Aspects, Pathology and Pathogenesis, Acta Odont. Scand. (Suppl.) 49:31-124 (1967). Eggen, K. and G. Rolla: Further Information on the Composition of the Acquired Pellicle. Cariology Today, Int. Congr., Zurich, pp. 109-112 (1983). Featherstone, J. D. B., J. M. Ten Cate, M. Shariati, and J. Arends: Comparison of Artificial Caries-Like Lesions by Quantitative Microradiography and Microhardness Profiles. Caries Res. 17:385-391 (1983). Hay, D. I. and E. C. Moreno: Statherin and the Acidic Proline-Rich Proteins. In: Human Saliva: Clinical Chemistry and Microbiology, Vol. 1, pp. 131-150. (J. 0. Tenovuo, Ed.) CRC Press Inc., Boca Raton (1989). Jensen, J. L., M. S. Lamkin, and F. G. Oppenheim: Adsorption of Human Salivary Proteins to Hydroxyapatite. J. Dent. Res. 71:1569-1576 (1992). Lie, T.: Pellicle Formation on Hydroxyapatite Splints Attached to the Human Dentition: Morphologic Confirmation of the Concept of Adsorption. Arch. Oral. Biol. 20:739-741 (1975). Mandel, I. D.: Relation of Saliva and Plaque to Caries. J. Dent. Res. 53:246-266 (1974). Mandel, I. D.: Dental Caries. American Scientist 67:680-688 (1979). Meckel, A.: The Formation and Properties of Organic Films on Teeth. Arch. Oral. Biol. 10:585-597 (1965). Ten Cate, J. M. and J. D. B. Featherstone: Mechanistic Aspects of the Interactions between Fluoride and Dental Enamel. Crit. Rev. Oral Biol. Med. 2:283-296 (1991). White, D. J.: Use of Synthetic Polymer Gels for Artificial Carious Lesion Preparation. Caries Res. 21:228-242 (1987). White, D. J. and J. D. B. Featherstone: A Longitudinal Microhardness Analysis of Fluoride Dentifrice Effects on Lesion Progression In Vitro. Caries Res. 21:502-512 (1987). Zahradnik, R., E. Moreno, and E. Burke: Effect of Salivary Pellicle on Enamel Subsurface Demineralization In Vitro. J. Dent. Res. 55:664-670 (1976). Zahradnik, R., D. Propas, and E. Moreno: In Vitro Enamel Demineralization by Streptococcus mutans in the Presence of Salivary Pellicle. J. Dent. Res. 56:1107-1110 (1977). Zahradnik, R., D. Propas, and E. Moreno: Effect of Salivary Pellicle Formation Time on In Vitro Attachment and Demineralization by Streptococcus mutans. J. Dent. Res. 57:1036-1042 (1978). Zahradnik, R.: Modification by Salivary Pellicles of In Vitro Enamel Remineralization. J. Dent. Res. 58:2066-2073 (1979).

362
Downloaded from cro.sagepub.com at UNIVERSITAS GADJAH MADA on December 24, 2013 For personal use only. No other uses without permission.