Serge Bouillaguet, DMD, DCD, John C. Wataha, DMD, PhD, Carl T. Hanks, DDS, PhD, ..... hance its cytotoxicity needs to be investigated, especially because.
0099-2399/96/2205-0244503.00/0 Printed in U.S.A. VOL. 22, NO. 5, MAY 1996
JOURNAL OF ENDODONTLCS
Copyright © 1996 by The American Association of Endodontists
In Vitro Cytotoxicity and Dentin Permeability of HEMA Serge Bouillaguet, DMD, DCD, John C. Wataha, DMD, PhD, Carl T. Hanks, DDS, PhD, Bernard Ciucchi, DMD, and Jacques Holz, DMD, PhD
eral different ways to regulate diffusion. Dentin is a barrier to convection and diffusion that reduces the concentrations of substances that reach the pulpal ends of the dentin tubules. According to the Fick equation, the rate of diffusion is dependent on the applied concentration, but inversely proportional to the dentin thickness. The surface area available for diffusion, the temperature, and the chemical characteristics of the diffusing molecules all affect diffusion (7). Furthermore, positive pulpal tissue pressure may also reduce the concentration of solutes in the diffusate (8). Pashley and Matthews (9) have recently reported a significant decrease in the rate of diffusion of radioactive iodine when an outward convective fluid movement is present. Finally, pulpal blood flow can remove toxic materials from the pulp once they diffuse through the dentin (10). Although the dynamics of the pulpodentinal complex cannot be reproduced under laboratory conditions, different 'in vitro pulp chambers' have been proposed for cytotoxicity testing (11). These studies have clearly demonstrated a reduced apparent cytotoxicity of solutes when the solutes are separated from cells by a dentin barrier. However, under these experimental conditions, materials that diffused through the dentin were diluted by a volume (up to 4 ml) of the cell culture medium on the pulpal side of diffusion chamber (12, 13). Thus, the exact relationship between the concentration of the difl'used solute and the observed cytotoxic response could not be established. In more recent studies (14), including the present study, an alternative strategy was used to identify the concentrations of a resin component that may cause cytotoxic responses if applied to dentin. First, the dose response of the cells to the component was determined in direct contact with cells. Second, the diffusion of the component across dentin was measured under various conditions. Using these two results, it was possible to identify the concentrations of the component that may cause cytotoxicity if applied to the dentin. Specifically, the objectives of the current study were to (a) determine a dose-response curve for HEMA in direct contact with cells in vitro and (b) measure the concentration of HEMA that diffused through dentin in vitro as a function of dentin thickness, hydraulic conductance, and presence or absence of simulated pulpal pressure.
An in vitro diffusion chamber was used to measure the diffusion of 2-hydroxyethyl methacrylate (HEMA) through etched human dentin disks. Concentrations of HEMA, which diffused through dentin, were measured by ultraviolet spectroscopy, and the effect of initial HEMA concentration, dentin thickness, and back pressure on diffusion were assessed. The cytotoxicity of HEMA was determined using BALB/c 3T3 mouse fibroblasts in direct contact with HEMA for 12 or 24 h. HEMA diffused rapidly through dentin under all conditions, but increased thickness, back pressure, or decreased initial concentration all reduced diffusion. The permeability coefficient of HEMA was -0.0003 cm/min, and diffusion through 0.5 mm of dentin reduced the HEMA concentration by a factor of - 6 , 0 0 0 (with 10 cm of H20 back pressure). It was concluded that the risk of acute cytotoxicity to HEMA through dentin was probably low, but that decreased dentin thickness, lack of polymerization, or extended exposure times might increase the risk significantly.
For many years, 2-hydroxyethyl methacrylate (HEMA) has been incorporated into many dentin bonding agents used in dentistry (1). HEMA contains both hydrophylic and hydrophobic groups, and can promote the diffusion of methacrylate-based monomers into the dentin after acidic treatment. This penetration results in the formation of a resin-infiltrated dentin zone called the hybrid layer (2) or resin-dentin interdiffusion zone (3). Although the use of resin monomers such as HEMA has been claimed to improve bond strength of composites to dentin (4), Pashley et al. (5) have recently reported the persistence of fluid-filled channels within this hybrid layer caused by incomplete impregnation. Thus, water-soluble components such as HEMA could easily diffuse through the dentin tubules and affect the underlying odontoblastic cell layer. Moreover, the low molecular weight of HEMA might allow a faster rate of diffusion across dentin, compared with other larger resin molecules that are relatively insoluble in aqueous solutions (6). The pulpodentinal complex can interact with molecules in sev-
MATERIALS AND METHODS
Serial dilutions of HEMA (batch 404194; Aldrich, Milwaukee, WI) were prepared in double-distilled water to obtain a standard 244
Vol. 22, No. 5, May 1996
Cytotoxicity and Permeability of HEMA
245
~ , 0.3
,.~ 0.2 FIG 2. Diagram of the apparatus used to measure the diffusion of HEMA through dentin. R, pressure reservoir; P, peristaltic pump.
~
0.1
0 0.0
I
I
0.5
1.0
,,
I 1.5
2.0
2.5
Optical Density (no units) FIG 1. Standard curve of the concentration of HEMA as a function of optical density at 215 nm. The relationship seemed linear up to 0.30 mmol/L. The detection limit was - 5 / x m o l / L (0.005 mmol/L), and the optical density was -0.05.
curve of optical density versus concentration of HEMA. The optical density of the H E M A solution was measured in a Beckman DU-64 spectrophotometer at 215 nm against double-distilled water (Beckman Instruments, Inc., Fullerton, CA). This wavelength was selected based on preliminary scans of H E M A solutions. The concentrations of H E M A ranged from 0.01 to 0.30 mmol/L, and were selected based on pilot studies that showed a linear relationship between optical density and concentrations (Fig. I). The cell culture testing was performed using BALB/c 3T3 mouse fibroblasts (ATCC CCL 163, clone A31; American Type Culture Collection, Rockville, MD). Cells were maintained in Dulbecco's Modified Eagle's Medium (Sigma Chemical Co., St. Louis, MO) supplemented with 3% NuSerum (Collaborative Research, Bedford, MA), 1/xg/ml of gentamycin (Flow Laboratories, Inc., McLean, VA), penicillin (125 units/ml), streptomycin (125 /xg/ml), and glutamine (2 mmol/L, Gibco, Gaithersburg, MD). Twenty-four hours before the experimental procedures, the cells were plated at 25,000 cells/cm z in 96-well cell culture dishes (Costar, Cambridge, MA) containing 200/xl of medium/well. Cells were incubated at 37°C in a humidified 5% CO2:95% air atmosphere. Ten concentrations of HEMA were tested, ranging from 0.01 to 100 mmolFL. These concentrations were selected from a pilot study to cover the range of concentrations of HEMA which were likely to diffuse through dentin. There were six replicates for each concentration of HEMA. Negative controls were prepared by adding phosphate-buffered saline (PBS) to the cell culture medium rather than the HEMA solution. After addition of the HEMA- or PBS-containing medium, the cells were incubated for 12 or 24 h, then assessed for succinyl dehydrogenase (SDH) activity in the mitochondria by the MTT colorimetric assay previously described in detail (15). Briefly, the SDH activity was measured by replacing the cell culture medium with MTT (Sigma) solution. For over 1 h, active mitochondria converted the MTT to its blue, insoluble formazan salt. Thus, the amount of blue stain that accumulated in the cells was proportional to the number of active mitochondria. The water-insoluble formazan was subsequently dissolved in dimethyl sulfoxide, and the resulting purple color was quantified in a spectrophotometer at
560 nm. The SDH activity for the HEMA-containing wells was expressed as a percentage of negative controls. The diffusion chamber used in this experiment was modified from the original "split chamber" device previously used to measure dentin diffusion and hydraulic conductance (16). The device held a dentin disk between two silicone rubber O-rings and was connected to a spectrophotometer (Fig. 2). Dentin disks were prepared from freshly extracted human third molars that were not carious. The teeth were cut in cross-section at the midlevel of the crown using a diamond-coated band saw (Exakt, Germany) under constant water coolant. Only one disk was taken from each tooth. Disks of 0.5- and 1.0-nun thickness were prepared and stored in 0.1% sodium azide to inhibit bacterial growth. Just before use, the disks were dipped for 2 min in 0.5 mol/L EDTA (pH 7.4) to remove the smear layer created during the cutting process. The disks were then rinsed with PBS for 3 min to stop chelation and were placed in the diffusion chamber. The surface area for diffusion was 0.29 cm 2, as defined by the diameter of the silicone O-rings. The total volume of the lower chamber, including the connecting tubing, was 70 txl. Hydraulic conductance was determined by connecting the chamber to a 180-cm column filled with double-distilled water. During this measurement, the outflow tubing was closed (Fig. 2), and fluid was permitted to flow through the dentin to displace air bubbles in the dentin tubules. The volume of fluid that flowed through the disk after 30 min was measured, and the hydraulic conductance was calculated (14). To maintain the position of the disk in the chamber, the disk was not removed from the chamber until all experiments were completed. Thus, a separate diffusion chamber was used for each disk. There were five replicates for each condition. For the diffusion studies, the chamber was connected directly to the spectrophotometer by opening the outflow tubing (Fig. 2). A constant flow of 0.2 ml/min of double-distilled water was moved through the chamber using a peristaltic pump (Buchler, Ft. Lee, N J). Considering that the action of this pump could have created a negative pressure inside the circuit, it has been calculated from the experimental parameters (i.e. fluid flow rate, tubing geometry, and fluid viscosity) that this set-up produced a pressure drop of 0.2 kPa inside the chamber. As reported by Pashley and Matthews (9) in a similar diffusion study, this pressure drop can be considered insignificant relative to the surrounding pressure (
