Comparison of in vitro erosion protocols in bovine teeth to simulate natural erosion lesion: analysis of mechanical properties and surface gloss

Abstract Objective The aim of this study was to compare two in vitro erosion protocols, in which one simulates in vivo conditions experienced by patients with gastroesophageal disorders or bulimia (HCl-pepsin protocol), and the other simulates the diet of an individual who consumes a high volume of erosive beverages (citric acid protocol). In addition, the mechanical properties and surface gloss of eroded human dentin were compared with those of sound human dentin. Materials and Methods Blocks of cervical dentin were used: sound human dentin (n=10), human dentin with erosive lesions (n=10), and bovine dentin (n=30). Twenty bovine blocks were subjected to either of two erosion protocols (n=10/protocol). In the first protocol, samples were demineralized using HCl-pepsin solution, then treated with trypsin solution. In the second protocol, samples were demineralized with 2% citric acid. Toothbrushing was performed in both protocols using a toothbrushing machine (15 s with a 150 g load). Ten bovine dentin blocks were not subjected to any erosive treatment. All samples of bovine and human dentin were analyzed to obtain Martens hardness values (MH), elastic modulus (Eit*) and surface gloss. One-way ANOVA and Tukey's test were performed to analyze the data (α=0.05). Results Sound human and eroded human dentin groups showed similar MH and Eit* values (p>0.05); however, sound human dentin showed a higher surface gloss value when compared to eroded human dentin (p<0.05). Sound bovine dentin and HCl-pepsin-treated bovine dentin treatments resulted in similar values for both MH and Eit* (p>0.05), but HCl-pepsin-treated bovine dentin and citric acid-treated bovine dentin resulted in lower surface gloss than sound bovine dentin (p<0.05). Conclusions The HCl-pepsin protocol modified bovine dentin properties that could be similar to those that occur on human dentin surfaces with erosive lesions.


Introduction
The occurrence of non-carious cervical lesions (NCCL) is a very common clinical situation, resulting from multifactorial etiology. 1,2,3 Establishing differentiation among NCCL types is difficult because most cases involve the association of abfraction, attrition, erosion, and abrasion. 4,5 Acids from gastroesophageal disorders, foods and drinks play key roles in the development of erosive lesions, which may cause irreversible loss of dental tissue. 1,2,6,7 Given this context, dentin is a complex tissue which undergoes constant change due to the network of tubules that extend from the pulp to the dentinoenamel junction; thus, being closely related to the pulp. 8,9 Dentin hypersensitivity is common in cases of erosion. 1 The erosion process starts in the peritubular dentin, which shows a greater degree of mineralization and involves dentinal tubules. 10 Following, demineralization of hydroxyapatite crystals occurs in the intertubular dentin, thus exposing the collagen fibrils of the organic matrix. 10 In addition, the erosive process can involve metalloproteinases, natural constituents of dentin that function in the degradation of collagen fibers, which may accentuate the erosion process by degrading the collagen matrix. 11 Some modifications occur in eroded dentin areas, such as changes in mechanical properties 12 and brightness. 13 Due to the high incidence of in vivo dental erosion, there is a great need for protocols and studies that accurately reproduce the erosive process in vitro, to facilitate the better understanding of dentin tissue. 4,14 However, simulating all complex conditions that occur in the oral environment is very difficult 4 . In addition, the use of enzymes or acids to cause in vitro erosion requires optimization to ensure that the outcomes have similar characteristics to in vivo lesions 4 . Some prior studies have used gastrointestinal tract enzymes to simulate gastroesophageal disorders. 6,7 Another in vitro protocol involves the use of citric acid, a key ingredient in some beverages, which are associated with tooth erosion. 15,16 However, the literature lacks in vitro protocols that accurately simulate clinical conditions of erosive tooth wear. Therefore, the objective of this study was to compare two in vitro erosion protocols, in which one simulates in vivo conditions experienced by bulimia patients (HCl-pepsin protocol), and the other simulates the diet of an individual who consumes high volumes of erosive beverages (citric acid protocol); these protocols reproduce alterations of human dentin subjected to natural erosive lesions. The second objective was to compare the mechanical properties and surface gloss of eroded human dentin and sound human dentin. Two null hypotheses were tested: (1) there would be no differences in mechanical properties and surface gloss between sound and eroded human dentin; (2) there would be no differences in mechanical properties and surface gloss of bovine dentin after exposure to two in vitro erosion protocols when compared to sound bovine dentin.

Sample preparation
Twenty human pre-molars and lower incisors [sound human dentin (n=10) and human dentin with erosive lesions (n=10)] were collected from patients at a private clinic; the clinician who performed the surgical procedures provided information regarding the type of lesion in each case. Moreover, the clinician was able to detect these lesions through anamnesis, in which the patients reported their eating habits (ingestion of acidic beverages) or gastroesophageal disorders. All patients provided written informed consent to donate their teeth for the study. The All blocks of human dentin (sound or eroded) and 10 blocks of bovine dentin were stored at 100% relative humidity for 2 weeks in an incubator (ECB-2. Adamo Products for Laboratory Ltda., Piracicaba, SP, Brazil) at 37°C. The other 20 bovine dentin blocks were treated with the erosion protocols, as described below.

Results
Sound and eroded human dentin showed similar MH and Eit* values (Table 1), with no statistically significant differences (p>0.05). The power analysis was 100% for MH and 29% for Eit*. However, surface gloss showed higher values for sound human dentin than for eroded human dentin (p<0.05). Table 1 summarizes all data regarding mechanical properties and surface gloss.
Sound bovine dentin was not statistically different from bovine dentin that were subjected to the HClpepsin protocol considering MH and Eit* values (p>0.05 for both). Bovine dentin that underwent the citric acid protocol showed the lowest MH and Eit* values when compared to both sound and HCl-pepsin-eroded bovine dentin (p<0.05). These values are summarized in Table   Groups MH Eit* Surface gloss Human dentin with NCCL  2. Sound bovine dentin showed higher surface gloss values than other groups (p<0.05, Table 2). Regarding surface gloss, no statistical differences were found between bovine dentin that were subjected to either of the erosion protocols (p>0.05).

Discussion
According to the results obtained in this study, MH and Eit* values for both sound and eroded human dentin showed no statistical differences (Table   1). This may be a result of reparative or sclerotic processes that occur in eroded human dentin. 10,19 After experiencing erosion, odontoblasts within dentinal tubules can start the formation of tubular dentin with a new morphology, 10,20 create minerals that occlude the dentinal tubules 10,19 and form atubular dentin. 10 These processes may contribute to the maintenance of the mechanical properties of eroded human dentin, in relation to sound human dentin. The demineralized surface layer was softened after the erosive challenge 21 and could be removed by mechanical abrasions, such as toothbrushing, 21,22 increasing the hardness values of eroded human dentin.
In general, sound human dentin showed higher MH and Eit* values than sound bovine dentin. This result is consistent with a previous study 23 that found greater hardness values in human dentin because bovine dentin has a higher density of dentinal tubules. 23,24 Despite this finding, some authors consider bovine dentin to be suitable for in vitro experiments, especially when combined erosion and abrasion are applied. 25 Furthermore, some studies found no significant differences between human dentin and demineralized and mineralized bovine dentin in conventional hardness tests. 26 When comparing the citric acid and HCl-pepsin protocols, the citric acid protocol was found to be more aggressive in terms of dentin mineral dissolution ( Table   2). Citric acid chelates with calcium ions and may dissolve the smear layer of dentinal tubules, 15 thus contributing to the reductions in mechanical properties found in this study (Table 2) Pepsin can degrade the organic matrix by up to 25% without influencing mineral loss. 6 Moreover, the remineralizing agent, consisting of a trypsin solution (also used for sample storage and slurry formation), is supersaturated when considering hydroxyapatite. 7 These facts may explain the modified reduction in mechanical properties of specimens that were subjected to the HCl-pepsin protocol when compared to the citric acid protocol, and may also explain the lack of difference from sound bovine dentin (Table 2).
Another important factor is that proteoglycans (PG) and glycosaminoglycan (GAG) form supramolecular aggregates that are interconnected with the collagen network; these can regulate the behavior of the extracellular matrix. 28 Proteoglycans may act in the regulation of osmotic and hydrostatic pressure, as well as in the elastic behavior of the adjacent dentinal tubules. 28   decrease in mechanical properties was observed for dentins that were subjected to the HCl-pepsin protocol, with no difference from sound bovine teeth (Table   2). In addition, the abrasive load used for brushing (150 g) may not be able to remove the demineralized organic matrix. A study of different abrasive loads found that a toothbrushing abrasion load of up to 4 N (approximately 400 g) was unable to remove the remaining organic matrix and protect the underlying dentin; therefore, it could not decrease the progression of the erosion process. 29 In vitro protocols do not expose the samples to salivary dynamics that occur in the oral cavity; 4 importantly, oral buffering dynamics that could minimize the mineral loss process. 10,23 Although saliva contains enzymes that can degrade the organic matrix and accentuate the erosion process, it contains neutralization mechanisms and promotes remineralization of the substrate. 11,30 Furthermore, in vivo erosion, which is strongly linked to the individual's lifestyle, is not continuous and standardized, as occurs in in vitro protocols. 10,23 To simulate erosive lesions more accurately the erosion should be similar to the native pattern found in human dentin. The bovine dentin that was subjected to the HCl-pepsin protocol was exposed to enzymes that are natively in contact with the oral cavity when vomiting, or even in cases of reflux. This erosion protocol resulted in the modification of dentin properties that could be similar to those that occur in eroded human surfaces.
The methods used in this study aimed to quantify the mechanical behavior of dentin in different situations. Using conventional micro-hardness tests would not have been adequate, since the substrate presents elasticity due to possible contraction of the exposed organic matrix, and other conditions could alter the micro-hardness values. 31  Eroded human dentin showed lower surface gloss than sound human dentin (Table 1), rejecting the first null hypothesis of this study. Sound bovine and human dentin groups were polished and not subjected to any erosive protocol; therefore, their surface gloss values represent the initial unaltered values. Both erosive protocols decreased the surface gloss of bovine dentin when compared to sound bovine teeth ( Table   2). Thus rejecting the second null hypothesis of this study. Ganss, et al. 10 (2014) showed that after an in vitro erosion protocol the specimens were softened and dull. This corroborates the results obtained in this study, which showed loss of surface gloss regardless of the erosive protocol employed. Both in vitro erosive protocols resulted in similar decreases in surface gloss values. Nonetheless, when considering the mechanical properties, the modifications in dentin caused by the HCl-pepsin protocol were closer to modifications that occur in natural lesions than changes caused by the citric acid protocol. Namely, the comparison of the mechanical properties (MH and Eit*) of sound human dentin and eroded human dentin showed that they were similar (Table 1), as is the case for sound and HCl-pepsin-treated bovine dentin (Table 2). Other studies, such as profilometry and scanning electron microscopy, should be performed to confirm these results.
Although baseline values are important to assure that all samples were similar to mineral content, the lack of baseline and ΔMH measurements can be considered a limitation of this study, since it was not possible to conduct baseline measurements in the human dentin with NCCL before the lesions had occurred.

Conclusion
Bovine dentin that were subjected to the HClpepsin protocol, which uses enzymes present in the gastrointestinal tract, showed similar mechanical properties (MH and Eit*) when compared to sound bovine teeth and a decrease in surface gloss values.
Similarly, eroded human dentin showed similar mechanical properties and less surface gloss when compared to sound human dentin.