Characterization of white spot lesions formed on human enamel under microcosm biofilm for different experimental periods

Abstract The initial characteristics of white spot lesion (WSLs), such as the degree of integrated mineral loss (ΔZ), depth and pattern of mineral distribution, have an impact on further demineralization and remineralization. However, these lesion parameters have not been evaluated in WSLs produced from microcosm biofilms. Objective: This study characterized artificial white spot lesions produced on human enamel under microcosm biofilm for different experimental periods. Methodology: In total, 100 human enamel specimens (4x4mm) were assigned to 5 distinct groups (n=20/group) differing according to the period of biofilm formation (2, 4, 6, 8 or 10 days). Microcosm biofilm was produced on the specimens from a mixture of human and McBain saliva at the first 8h. Enamel samples were then exposed to McBain saliva containing 0.2% sucrose. WSLs formed were characterized by quantitative light-induced fluorescence (QLF) and transverse microradiography (TMR). Data were analyzed by ANOVA/Tukey or Kruskal-Wallis/Dunn tests (p<0.05). Results: A clear time-response pattern was observed for both analyses, but TMR was able to better discriminate among the lesions. Regarding QLF analysis, median (95%CI; %) changes in fluorescence ∆Z were -7.74(-7.74:-6.45)a, -8.52(-8.75:-8.00)ab, -9.17(-10.00:-8.71)bc, -9.58(-10.53:-8.99)bc and -10.01(-11.44:-9.72)c for 2, 4, 6, 8, and 10 days, respectively. For TMR, median (95%CI; vol%.µm) ∆Z were 1410(1299-1479)a, 2420(2327-2604)ab, 2775(2573-2899)bc, 3305(3192-3406)cd and 4330(3972-4465)d, whereas mean (SD; µm) lesion depth were 53.7(12.3)a, 71.4(12.0)a, 103.8(24.8)b, 130.5(27.2)bc, 167.2(39.3)c for 2, 4, 6, 8 and 10 days, respectively. Conclusion: The progression of WSLs formed on human enamel under microcosm biofilm can be characterized over 2-10 days, both by QLF and TMR analyses, although the latter provides better discrimination among the lesions.


Introduction
Worldwide, dental caries is the most prevalent chronic disease, being considered an important public health problem. 1 The lesion results from the development of a cariogenic biofilm due to the interaction, along time, of frequent sugar consumption, poor oral hygiene, and unfavourable host factors. 2,3 The first clinical sign of dental caries is a white spot lesion (WSL), that is characterized histologically by a subsurface demineralization below a pseudo-intact outermost enamel layer. At this stage, the lesion can still be remineralized. For this reason, several methodologies have been proposed to create artificial WSLs to delevop and evaluate new therapeutic approaches.
While for natural WSLs the presence of a biofilm is essential to develop the lesion, artificial lesions can be created using abiotic models, involving demineralizing agents or pH-cycling protocols 4 . Biotic in vitro models include single species, multi-species or a microcosm approach. 5 Microcosm biofilms that originate from the whole-mixed natural microbiota have the important advantage of representing the natural microbiota in its entirety, allowing the replication of the complex interactions within the oral ecosystem. 6,7 Moreover, microcosm biofilms are not steady-state systems, e.g. they evolve, thus resembling dental plaque in vivo. 5 Artificial WSLs are typically produced to test the efficacy of anti-caries products. 4 Thus, the initial characteristics of the lesions, such as the degree of integrated mineral loss (ΔZ), depth and pattern of mineral distribution, have an impact on further demineralization and remineralization, [8][9][10][11] which can affect the performance of the product that is being tested. However, these lesion parameters have not been evaluated in WSLs produced from microcosm biofilms. We raised the hypothesis that, by varying the experimental time, it is possible to produce distinct lesions in terms of mineral loss and depth, with different abilities to respond to the action of distinct remineralizing products. Thus, this study aimed to characterize artificial WSLs produced on human enamel under microcosm biofilm for different experimental periods.

Methodology Study design
The protocol of this study was approved by the local ethical committee (CAAE 99086718.0.0000.5417).
Saliva was collected from ten volunteers. In total, 100 enamel specimens were obtained from unerupted third molars and divided into 5 groups, according with the period of biofilm formation (2,4,6,8 or 10 days).
Microcosm biofilm was produced on the specimens from a mixture of human and McBain saliva at the first 8 h. Enamel samples were then exposed to McBain saliva containing 0.2% sucrose. WSLs formed were characterized by quantitative light-induced fluorescence (QLF) and transverse microradiography (TMR) (Figure 1).

Saliva collection
After informed consent was provided, saliva was collected from 10 healthy donors (22-35 years old) who met the inclusion criteria: (1) normal salivary flow (stimulated salivary flow > 1 ml/min and nonstimulated salivary flow > 0.3 ml/min), (2) caries history but no active caries (no active white-spot and/or cavitated lesions); (3) absence of gingivitis/ periodontitis (gum bleeding or tooth mobility); and (4) no ingestion of antibiotics for three months prior to the experiment. The exclusion criteria were the conditions opposite to those showed above, as well as individuals with chronic systemic diseases, smokers, pregnant, and lactating women. Prior to the day of collection, the donors did not brush their teeth. Furthermore, they were not allowed to ingest food or drinks within the last 2h before saliva collection. The saliva was collected under stimulation by chewing a rubber material for 10 min in the morning. After collection, saliva was pooled and diluted in glycerol (70% saliva and 30% glycerol).

Microcosm biofilm formation
The human saliva was defrosted and mixed with

The solution of human saliva and McBain saliva
was added to each well containing an enamel sample (v=1.5 ml well −1 ) in a 24-well plate, which was incubated at 5% CO 2 and 37°C. After 8 h, the medium was removed, the enamel samples were washed using PBS (Phosphate Buffered Saline) (5s) and fresh McBain saliva now containing 0.2% sucrose was added to the wells (v=1.5 ml well −1 ). The microplates were incubated at 5% CO 2 and 37°C for another 16h, completing the first day. Every 24 h, the medium was changed and incubated in the same conditions as described above until completing 2, 4, 6, 8 or 10 days of culture 12,16 ( Figure 1).

Quantitative Light-Induced Fluorescence
QLF was applied to measure the changes in the enamel fluorescence. A xenon arc lamp was used as a light source, and an optical filter system, producing blue light with a maximum wavelength of 370 nm, was Electric Industrial Co, Ltd, Osaka, Japan) equipped with high pass yellow filter (γ>520 nm) to exclude any excitation or ambient light that might reach the detector, and a special dental mirror to reflect the image of the lesion connected to the camera.
After drying the sample surface (for 5s), images were obtained by QLF, in a completely dark environment.   p<0.0001). Median ∆Z for 2 days was significantly lower than that obtained for 6, 8 and 10 days. Median ∆Z for 4 days was significantly lower than that found for 8 and 10 days and also for 6 days was lower than for 10 days. The other differences were not significant.
Regarding lesion depth, ANOVA found a significant difference among the groups (F=42.892, p<0.0001).
Lesions formed for 2 and 4 days had similar depths that were significantly lower than the ones found for the formed from 6 days on. Lesions formed for 6 days had depth values significantly lower than those of the lesions formed for 10 days. The other differences were not significant. The pattern of mineral distribution within the lesions formed in the different experimental periods was similar and no significant differences among the mean R values were detected (F=0.865, p=0.491) ( Table 1). Figure 2 shows the representative TMR radiographs of the lesions formed in each experimental period.

Discussion
In our study, we employed a microcosm biofilm model to produce WSLs with different degrees of mineral loss and depth, based on the number of experimental days to which the enamel specimens were subejcted to the microcosm biofilms. This type of biofilm represents the natural microbiota in its entirety, 20 offering the advantage of replicating the complex interactions within the oral ecosystem. 6,7 Furthermore, its advantage compared to other in vitro caries models, such as pH-cycling model, is that we can study the antibacterial effect of anticaries agents.
Then, we can consider microcosm biofilm as a preclinical model. 21,22 In the model, 0.2% sucrose was continuously available in McBain´s artificial saliva from 8 h of the beginning of the experiment. 15 This procedure was different from other models that employed shorter periods of exposure to higher sucrose concentrations, 23-25 in which microcosm biofilm was grown under continuous 0.2% sucrose exposure.
We evaluated intermittent exposures to 1% sucrose; however, we were unable to induce dental caries lesion formation in 5 days of biofilm growth. Therefore, we kept the continue exposure to sucrose, although this approach does not simulate the in vivo intermittent exposure to sugar from a regular diet. Although the lesion in human enamel was lower than those induced in bovine enamel, it was detectable by two methods of analysis: QLF (clinical analysis) and TMR (laboratory analysis). As expected, TMR was more sensitive to detect changes in mineral content and lesion depth than QLF. In fact, TMR is regarded as the gold standard technique to evaluate mineral content, while QLF has the advantage of being employed clinically. However, this study did not present clinical characteristics, so this advantage was not important. the culture medium was adjusted to 4.5 or 7 (during demineralization and remineralization, respectively).
The authors did not find a time-response pattern regarding ∆Z and lesion depth 23 as we did. In our study, R showed no change in the different experimental periods, which means that the mineral losses were proportional along to the depths. Therefore, the increase of ∆Z was due to the progression of the lesion to deeper layers.
The degree of mineral loss and lesion depth is known to play an important role in mineral diffusion. 8,29 Lesions with higher ΔZ at baseline have a pronounced In addition to the initial mineral loss of WSLs, its mineral distribution is also important. 4 In our study, the R values were similar among the groups, which is expected since in vitro R is constant over the demineralization period. 33 Moreover, low-R lesions tend to be more suitable when physiological mineral distribution is required. On the other hand, high-R lesions can better discriminate among the treatments being studied. 8 The R of natural enamel WSLs has been reported to be around 16,8 lower than the values found in the present study, which ranged between 25 and 30, as expected, since the natural lesions are usually developed over a longer period of time.
This means that the lesions formed under microcosm biofilm, according with our protocol, might be more appropriate to distinguish among different treatments.
Considering these aspects, to compare results from studies having lesions with distinct characteristics at baseline requires caution, 4

Funding
This study was supported by LAOHA (Latin America Oral Health Association).

Conflicts of interest
There are no conflicts of interest.