|Home | About | Journals | Submit | Contact Us | Français|
Matrix metalloproteinases (MMPs) cause collagen degradation in hybrid layers created by dentin adhesives. This in vitro study evaluated the feasibility of using a cross-linking agent, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC), to inactivate soluble rhMMP-9, as an example of dentin MMPs, and matrix-bound dentin proteases. The inhibitory effects of 5 EDC concentrations (0.01-0.3 M) and 5 incubation times (1-30 min) on soluble rhMMP-9 were screened with an MMP assay kit. The same EDC concentrations were used to evaluate their inhibitory effects on endogenous proteinases from completely demineralized dentin beams that were incubated in simulated body fluid for 30 days. Decreases in modulus of elasticity (E) and dry mass of the beams, and increases in hydroxyproline content of hydrolysates derived from the incubation medium were used as indirect measures of matrix collagen hydrolysis. All EDC concentrations and pre-treatment times inactivated MMP-9 by 98% to 100% (p < 0.05) compared with non-cross-linked controls. Dentin beams incubated in 0.3 M EDC showed only a 9% decrease in E (45% decrease in control), a 3.6% to 5% loss of dry mass (18% loss in control), and significantly less solubilized hydroxyproline when compared with the control without EDC cross-linking (p < 0.05). It is concluded that EDC application for 1 min may be a clinically relevant and effective means for inactivating soluble rhMMP-9 and matrix-bound dentin proteinases if further studies demonstrate that EDC is not toxic to pulpal tissues.
Etch-and-rinse dentin adhesives rely predominantly on micromechanical retention for bonding resins to dentin. Acid-etching of dentin exposes matrix metalloproteinases (MMPs) which are inactive in mineralized dentin and bone (Collins et al., 2002; Breschi et al., 2010). These enzymes are zinc- and calcium-dependent hydrolases that add water across specific peptide linkages in collagen peptides (Nagase et al., 2006). Upon activation by mildly acidic adhesive resin components (Mazzoni et al., 2006; Tay et al., 2006), MMPs (MMP-2, -8, -9) (Martin-De Las Heras, 2000; Sulkala et al., 2002, 2007; Mazzoni et al., 2006) cause progressive degradation of exposed collagen fibrils.
Chlorhexidine has been used as a non-specific MMP inhibitor (Gendron et al., 1999) to prevent degradation of hybrid layers (Hebling et al., 2005; Carrilho et al., 2007; WW Brackett et al., 2007; MG Brackett et al., 2009). However, chlorhexidine is water-soluble and may leach out of hybrid layers, compromising its long-term anti-MMP effectiveness (Ricci et al., 2010). An entirely different approach is to treat the acid-etched dentin containing activated matrix-bound MMPs with cross-linking agents that inactivate the catalytic site of proteases (Liu et al., 2011). Carbodiimides have been used as alternative cross-linking agents to gluteraldehyde, since they contain no potentially cytotoxic aldehyde residuals (Huang et al., 1990; Petite et al., 1995). They non-specifically cross-link all proteins by activating the carboxylic acid groups of glutamic and aspartic acids to form an O-acylisourea intermediate. The latter reacts with the ϵ-amino groups of lysine or hydroxylysine to form an amide cross-link, leaving urea as the terminal by-product. Carbodiimides are examples of zero-length cross-linking agents wherein peptides are cross-linked to one another without introducing additional linking groups. Previous research utilized EDC to increase the durability of resin-dentin bonds by increasing the mechanical properties of the collagen matrix (Bedran-Russo et al., 2010); however, the 1 to 4 hrs required for that procedure is clinically unacceptable.
Calero et al. (2002) compared the effectiveness of glutaraldehyde and EDC in inactivating MMPs in porcine pericardium after chloroform-methanol extraction of lipids. In that study, 0.3 M EDC coupled with 0.05 M N-hydroxysuccinimide inhibited both MMP-2 and MMP-9 by 80%, whereas glutaraldehyde inhibited MMP-2 only by less than 25%. Thus, the objective of the present study was to evaluate the feasibility of using EDC to inactivate soluble rhMMP-9 as a representative dentin MMP, and in inactivating the endogenous collagenolytic activity of demineralized dentin. The null hypotheses tested were: (1) that EDC does not inactivate soluble rhMMP-9, (2) that EDC does not inactivate matrix-bound MMPs, and (3) that EDC does not inactivate matrix-bound MMPs within 1 min of treatment.
Purified rhMMP-9 and a generic colorimetric MMP assay kit (Sensolyte) were obtained from AnaSpec Inc. (Freemont, CA, USA). To evaluate the effect of EDC incubation time on inactivation of a representative soluble MMP, we dissolved EDC in pH 7 assay buffer to reach a final concentration of 0.3 M. Two different experiments were performed. The first was to determine the concentration of EDC required to inactivate rhMMP-9 completely. The enzyme was pre-incubated in each EDC concentration for 20 min prior to the assay of its activity. The second experiment was to determine how rapidly 0.3 M EDC could inactivate rhMMP-9.
To evaluate the effect of EDC concentration on MMP-9 inactivation, we diluted 0.3 M EDC with de-ionized water to 0.01, 0.02, 0.05, or 0.1 M. After pre-incubation of EDC with rhMMP-9 for 20 min, additional assay buffer and substrate were added (see Appendix). A proprietary MMP kit inhibitor (GM6001) from the assay kit was used as the control. Details of the experiments can be found in the Appendix.
Since the normality and homoscedasticity assumptions of the data were valid, % inactivation was analyzed by two-way ANOVA (concentration vs. time) and Tukey multiple-comparison tests at α = 0.05.
Sixty extracted human third molars were obtained from 18- to 21-year-old patients under a protocol approved by the Georgia Health Sciences University. The teeth were stored at 4ºC in 0.9% NaCl containing 0.02% NaN3 to prevent bacterial growth. The enamel and superficial dentin of each tooth were removed by means of an Isomet saw (Buehler Ltd., Lake Bluff, IL, USA) under water cooling. Dentin beams with dimensions 6 x 2 x 1 mm were sectioned from the mid-coronal dentin (60 beams).
The beams were completely demineralized in 10 wt% H3PO4 (pH 1) for 18 hrs at 25°C. Digital radiography was used to confirm the absence of residual minerals. The initial modulus of elasticity of each demineralized beam was determined with three-point bending. Ten beams were assigned to each of 6 groups (N = 10) so that the mean initial modulus of elasticity of each group was statistically similar. Five EDC concentrations (0.01, 0.02, 0.05, 0.1, or 0.3 M) were evaluated as previously described. Demineralized dentin beams from each experimental group were dipped into an EDC solution of the designated concentration for 1 min and blot-dried and then dropped into 1 mL of medium. The control group consisted of dentin beams without EDC pre-treatment. Each beam was in 1 mL of a calcium- and zinc-containing simulated body fluid (SBF) in labeled polypropylene tubes. This diluted the EDC concentration in the final medium below cross-linking values. The SBF contained 5 mM HEPES, 2.5 mM CaCl2.H2O, 0.05 mM ZnCl2, and 0.3 mM NaN3 (pH 7.4). The sealed tubes were incubated in a shaker-water bath (60 cycles/min) at 37°C for 30 days.
Each dentin beam was placed over a miniature three-point bending device with a span length of 2.5 mm (Carrilho et al., 2009; Tezvergil-Mutluay et al., 2010, 2011). A universal testing machine with a 1 N load cell (Transducer Techniques, Temecula, CA, USA) was used to load the beams to 10% strain while immersed in de-ionized water with a displacement rate of 0.5 mm/min. After maximum displacement, the load was returned immediately to 0% stress without further holding to prevent creep of the demineralized collagen matrix. Following initial testing, each beam was returned to its designated polypropylene tube. After 30 days of incubation, the beams were rinsed free of salts for 10 min with running distilled water and re-tested with three-point bending under the same conditions.
Stress-strain curves were prepared from the load-displacement data. The modulus of elasticity (E; in MPa) of each specimen was calculated as the slope of the linear portion of the stress-strain curve (see Appendix). We analyzed the data by a two-factor repeated-measures analysis of variance to examine the effects of EDC concentrations (i.e., control – 0.01 M, 0.02 M, 0.05 M, 0.1 M, 0.3 M EDC) and the repeated-factor storage time (i.e., baseline vs. 30 days) and the interaction of these two factors on the modulus of elasticity. All pair-wise multiple comparisons were performed by the Tukey test. Statistical significance was set at α = 0.05.
Loss of dry mass after the 30-day incubation period was used as an indirect measure of MMP-induced hydrolysis of the collagenous dentin matrix (Carrilho et al., 2009; see Appendix). Since the normality and homoscedasticity assumptions of the data were violated, loss of dry mass from the demineralized beams was analyzed with Kruskal-Wallis one-way ANOVA and Dunn’s multiple-comparison tests at α = 0.05.
Demineralization of the mineralized collagen matrix exposes and activates endogenous MMPs, even though they remain bound to the collagen (Dayan et al., 1983; Martin-De Las Heras et al., 2000). We determined the third index of matrix degradation by measuring the amount of solubilized collagen peptides over the 30-day incubation period (Carrilho et al., 2009). A hydroxyproline assay (Jamall et al., 1981; see Appendix) was used for estimation of the percentage of degraded collagen. For each specimen, the solubilized collagen was expressed as µg of hydroxyproline/mg of the dry mass of demineralized dentin before incubation. The data were evaluated by Kruskal-Wallis one-way ANOVA and Dunn’s multiple-comparison tests at α = 0.05.
The effect of incubation time of 0.3 M EDC on inhibition of the enzyme activity of soluble rhMMP-9 varied from 98% to 101% (Fig. 1B). There was no significant difference in the inactivation of rhMMP-9 between 1 min and 30 min of incubation (p > 0.05) (Fig. 1B). The inactivation effects of 20 min of pre-incubation of the 5 EDC concentrations on rhMMP-9 varied between 97% and 100% (Fig. 1A). Significant differences were observed among the 6 groups (p = 0.004); however, there were no statistically significant differences among the 5 EDC concentrations in inhibiting rhMMP-9 (p > 0.05).
The initial mean E of the demineralized dentin beams was 2.9 ± 0.7 MPa (Fig. 2). After 30 days of incubation, the control specimens in the simulated body fluid (SBF) lost 46% of their initial stiffness (Fig. 2). Dentin beams pre-treated with 0.3 M EDC lost only 9% of stiffness, whereas the 0.01 M EDC pre-treated beams lost 30% of their stiffness. EDC concentration had a significant effect on the E of the experimental beams (p < 0.05). There were significant interactions between EDC concentrations and time (p < 0.05).
Loss of dry mass (Fig. 3) of experimental beams over the 30-day incubation period showed significant differences from the SBF controls. Dentin beams incubated with EDC showed 3.5% to 5% decrease in dry mass, compared with an 18% decrease in the SBF control (p < 0.05). Among the EDC groups, a slight dose-response was observed; however, the difference was not significant (p > 0.05), except for the lowest EDC concentration (0.01 M; p < 0.05). The group with the highest EDC concentration (0.3 M) showed a dry mass loss of 3.6 %, whereas the 0.01 M pre-treated group lost around 5%.
Similar to the loss of dry mass and elastic modulus measurements, an EDC dose-response was observed with the solubilization of collagen peptides (Fig. 4). Control beams stored in the SBF medium liberated 9.4 µg hydroxyproline/mg dentin, whereas beams pre-treated with various concentrations of EDC liberated between 0.74 and 1.5 µg hydroxyproline/mg dentin. Beams pre-treated with the highest EDC concentrations (0.1 and 0.3 M) showed lower degrees (p < 0.05) of collagen solubilization compared with the beams pre-treated with the lowest (i.e., 0.01 M) EDC concentrations.
Carbodiimide has previously been investigated as a cross-linking agent to increase the mechanical properties of the dentin matrix (Bedran-Russo et al., 2010). However, a 1- to 4-hour application time was required for the mechanical properties of the collagen matrix to increase, making it impossible for use in clinical applications (Bedran-Russo et al., 2010). The results of the present study indicate that even a short pre-treatment (i.e., 1 min) of acid-etched dentin matrix is sufficient to inactivate endogenous protease activity of dentin without significantly stiffening the collagen.
Since the rhMMP-9 inactivation potential of non-specific EDC was similar to or higher than that of the specific MMP inhibitor GM6001, the first null hypothesis—that EDC does not inhibit soluble rhMMP-9—was rejected. The results of the soluble rhMMP-9 assay showed that even at 0.01 M, EDC successfully inhibited rhMMP-9 by over 99%. This is in line with previous work which showed a large decrease in MMP-9 activity following treatment with 0.1 M EDC for 4 hrs (Calero et al., 2002).
Based on the rhMMP-9 results, a 1-minute pre-treatment time was selected for inactivation of matrix-bound protease activity in completely demineralized dentin. Previous work has shown that endogenous proteases in dentin matrices slowly lower the modulus of elasticity of incubation time as collagen peptides are slowly solubilized (Carrilho et al., 2009; Tezvergil-Mutluay et al., 2010, 2011). When the decreases in stiffness of various groups were compared, the highest residual stiffness was obtained in the group treated with 0.3 M EDC, whereas the lowest stiffness was observed with the control (46% drop), where no pre-treatment was used. This suggests that the endogenous protease activities in the EDC pre-treated dentin beams were much lower than those in the control beams. The same trend was observed with the evaluation of dry mass and the hydroxyproline assay, which are indirect measures of solubilization of collagen from the dentin matrix. The loss of dry mass in the control group was around 5 times more than in the EDC pre-treated groups. Likewise, the control group released 5 to 6 times more hydroxyproline compared with the EDC pre-treated groups. This requires rejection of the second null hypothesis, that EDC does not inactivate matrix-bound proteases. We speculate that the EDC cross-links peptide chains in MMPs and cysteine cathepsins (Tersariol et al., 2010; Nascimento et al., 2011) of the dentin matrix. Cross-linking decreases the molecular mobility of the catalytic sites in these enzymes, which are critical for their function (Liu et al., 2011). This cross-linking is done very rapidly in acid-etched dentin that is 30 vol% collagen and 70 vol% water (Pashley et al., 2011). The carboxyl and amino groups in collagen may not be as accessible as those in MMPs (Orgel et al., 2006; Perumal et al., 2008), thereby permitting more rapid cross-linking of MMPs than of collagen.
There are potential advantages of inactivating the proteolytic enzymes of the dentin matrix (i.e., MMPs and cathepsins) by using EDC instead of using non-specific MMP inhibitors such as chlorhexidine, or specific MMP inhibitors like Galardin (Breschi et al., 2010). To be effective, enzyme inhibitors must remain firmly bound to the enzyme. While specific MMP inhibitors such as Galardin can inhibit MMPs at micromolar concentrations (Galardy et al., 1994), chlorhexidine requires millimolar concentrations, because collagen is not saturated by chlorhexidine until it reaches 30 mM (Kim et al., 2010). High concentrations of chlorhexidine (e.g., 2 wt%) can interfere with conversion of monomers to polymers (Hiraishi et al., 2010). Water-soluble inhibitors may slowly leach from resin-dentin interfaces over time. Recent studies showed that chlorhexidine inhibition of resin-dentin bonds was effective for 9 to 12 mos, after which it begins to lose its effectiveness (Ricci et al., 2010; Sadek et al., 2010). We speculate that functional stresses on hybrid layers may cause them to compress 0.6% under function and then rebound (Wood et al., 2008), which may allow unbound water within the hybrid layers to elute residual chlorhexidine from the acid-etched collagen matrix.
Within the limits of this in vitro study, it may be concluded that EDC can inactivate matrix-bound dentin proteinases in demineralized dentin matrices with a 1-minute application time. This result requires rejection of the third null hypothesis. Once cross-linked, these covalent bonds may be stable and therefore contribute to improved durability of resin-dentin bonds (Bedran-Russo et al., 2010) if newly synthesized odontoblast MMPs (Lehmann et al., 2009) in dentinal fluid cannot diffuse around resin tags in dentinal tubules bonded to dentin to reach the hybrid layer. This speculation needs to be tested by longitudinal in vivo studies and in vitro bond strength studies on the use of EDC to stabilize resin-bonded dentin, if EDC can be shown to be non-toxic to pulpal soft tissues. The use of EDC on acid-etched dentin would not prevent hydrolysis of the resin polymer component of the hybrid layer.
This work was supported by R01 DE015306-06 from the National Institute of Dental and Craniofacial Research (NIDCR) to DHP (PI) and by grant #8126472 from the Academy of Finland to AT-M (PI).
The authors thank Mrs. Michelle Barnes for her secretarial support. The authors do not have a financial interest in the products, equipment, and companies cited in the manuscript.
A supplemental appendix to this article is published electronically only at http://jdr.sagepub.com/supplemental.