In vivo model for microbial invasion of tooth root dentinal tubules

ABSTRACT

Objective

Bacterial penetration of dentinal tubules via exposed dentine can lead to root caries and promote infections of the pulp and root canal system. The aim of this work was to develop a new experimental model for studying bacterial invasion of dentinal tubules within the human oral cavity.

Material and Methods

Sections of human root dentine were mounted into lower oral appliances that were worn by four human subjects for 15 d. Roots were then fixed, sectioned, stained and examined microscopically for evidence of bacterial invasion. Levels of invasion were expressed as Tubule Invasion Factor (TIF). DNA was extracted from root samples, subjected to polymerase chain reaction amplification of 16S rRNA genes, and invading bacteria were identified by comparison of sequences with GenBank database.

Results

All root dentine samples with patent tubules showed evidence of bacterial cell invasion (TIF value range from 5.7 to 9.0) to depths of 200 mm or more. A spectrum of Gram-positive and Gram-negative cell morphotypes were visualized, and molecular typing identified species of Granulicatella, Streptococcus, Klebsiella, Enterobacter, Acinetobacter, and Pseudomonas as dentinal tubule residents.

Conclusion

A novel in vivo model is described, which provides for human root dentine to be efficiently infected by oral microorganisms. A range of bacteria were able to initially invade dentinal tubules within exposed dentine. The model will be useful for testing the effectiveness of antiseptics, irrigants, and potential tubule occluding agents in preventing bacterial invasion of dentine.

Dentine; Root caries; Microbiology; Biofilms

INTRODUCTION

Root caries, pulpitis, and dentine hypersensitivity are becoming increasingly more problematic as the dentate human population ages¹⁷. Gingival recession can lead to exposure of dentine and tooth wear can result in opening of dentinal tubules on the exposed surface. When dentine becomes exposed as a result of gingival recession, or through dental caries, cracks, or microleakage around restorations, microorganisms are able to gain access to the tubules¹⁵. More microorganisms are found in the dentine adjacent to periodontal pockets than in healthy radicular dentine, and more bacteria are found in superficial root dentine than in middle dentine¹. Bacteria can also laterally invade the root surface along the incremental lines of cementum and then infiltrate the dentine²¹. Bacteria can penetrate through hypomineralized enamel into the dentine and contribute to pulpal pain symptoms of teeth with molar incisor hypomineralization¹⁰.

Bacterial persistence within the dentinal tubules may exacerbate development of root caries, and the formation of complex microbial communities within deeper dentine or root canal space²⁶ may play a significant part in endodontic treatment failure¹⁸. Evidence suggests that the bacteria that initially invade dentinal tubules are often from the genera Enterococcus and Streptococcus¹⁵. Enterococci in particular readily penetrate dentinal tubules^13,18. Root dentinal tubule invasion models in vitro have been utilized widely to study different bacterial penetration capabilities¹⁴ and model pulpal infections¹⁵, and to test the effects of antimicrobials^3,4 and irrigants¹². The models have provided valuable information on the mechanisms involved in growth and penetration of dentine⁵, and the potential for various agents, such as photodynamic therapy²⁹ to help with controlling infection. However, the various laboratory models usually incorporate dentine samples that are exposed to an artificial nutrient environment in order to achieve infection with relevant microorganisms. Under natural conditions, dentine would be exposed to salivary components, gingival fluid, immune system molecules, and potentially hundreds of different microorganisms¹⁹. Most in vitro dentine infection models employ conditions that are quite different from the natural in vivo infection environment.

Dentine studies in vitro have also included testing various compounds for ability to occlude tubules as desensitizing agents². Valuable information has been obtained about the properties and effectiveness of such agents²⁷, but these are only just beginning to be tested under suitable in vivo conditions. For example, West, et al.²⁸ (2011) determined the abilities of desensitizing toothpaste technologies to occlude patent dentinal tubules in a clinical environment. Healthy subjects wore lower intraoral appliances retaining dentine samples, and these were analyzed after 4 d of treatment for degree of occlusion²⁸. We have utilized the basis of that study to develop a model for microorganism invasion of dentinal tubules in vivo. This will provide a suitable platform by which to investigate bacterial invasion of dentine within a clinical environment, and to test for effectiveness of tubule-occluding or antimicrobial agents to prevent bacterial invasion of dentine.

MATERIAL AND METHODS

Root dentine

Non-carious, unrestored human canine or pre-molar teeth with single root canals were obtained from orthodontic extractions. Teeth were obtained with informed consent and the study was approved by Central and South Bristol Ethics Committee (REC ref. 04/Q2006/50). Following extraction, teeth were soaked in 2% sodium hypochlorite (NaOCl) for 48 h and any soft tissue remaining was removed. Prior to sectioning, roots were washed in copious amounts of water, and rinsing was repeated following sectioning to ensure no traces of NaOCl remained. Teeth were stored in sterile distilled H₂O at 4°C until required. Roots were sectioned using a water cooled steel bladed cutting machine (Isomet Saw, Buhler Ltd., Evanston, IL, USA). In brief, the crown and root tip were removed, the remaining root was cut into 0.5 cm lengths, and the cervical segments were longitudinally sectioned in such a way that the root canal was exposed. The root sections were then autoclaved (121°C, 20 min) in distilled H₂O, which did not visually affect tubule structures⁵, and stored at 4°C.

Preparation of intra-oral appliances

For each subject, a lower alginate impression was recorded in a perforated stock tray. Within 30 min the impressions were poured in Kaffir D dental stone and subsequently two lower-oral appliances were constructed from Forestacryl^® self curing acrylic (Pearson Dental Supply Co., Sylmar, CA, USA). Adams cribs were constructed to fit the mandibular first molars to aid retention and wire loops were constructed in an anterior and posterior trench region to hold the dentine samples in place (Figure 1). The cervical region root sections were mounted into the appliances in such a way that the buccal facing surface was flush with, or just below, the level of the surrounding acrylic surface. Before placement in the appliance, the root sections were dipped in sterile distilled water, the face to be in contact with the appliance was dried and the sample mounted onto a small drop of molten sticky wax within the trench of the appliance. Once all four sections were in place they were then further secured in position with the wire loop that was built into the appliance (Figure 1). The appliances were stored overnight at 4°C in a sterile airtight container containing damp tissue to prevent them drying out.

Figure 1
View of dental appliance in place on a dental mould (A) and a close-up view showing positioning of dentine samples (B). This shows the lower right appliance and the dentine pieces with pulpal faces outwards labelled A-D. A similar appliance was placed on the lower left, with four dentine pieces designated E-H. The dentine samples were retained with wax on the base side and with a metal retaining wire on the exposed side (arrowed in panel A). Appliances were custom made for each individual human subject

Experimental design

Ethical approval for this work was obtained from the Institutional Ethical Review Board (REC ref. 04/Q2006/50). Inclusion criteria were healthy volunteers aged 18 or over that could accommodate a lower buccal appliance. Exclusion criteria were: pregnancy, lactation, gross caries, unstable periodontal disease, antimicrobial medication within 7 d previously, orthodontic appliances that would interfere with the study evaluations, and tongue or lip piercing. Four subjects, age range 20-26 years, with informed consent, were fitted with the appliances which were then worn over the course of the next 15 d between 09:00 and 21:00 h. They were removed for 1 h twice a day for mealtimes and also for the period over which any drink other than water was consumed. At 21:00 h, subjects removed the appliances, brushed them with tap water, rinsed them in running water for 20 s, and then stored them overnight in an airtight container. Subjects brushed their teeth morning and evening with fluoridated dentifrice. On completion of the trial on day 15, root samples were removed from the appliances at 15:00 h. There were no adverse events.

Microscopic analysis of bacterial invasion

Six pieces of dentine from the appliances were fixed in 10% neutral buffered formalin for 7 d before being demineralized in 10% formic acid containing 2% formalin for 7 d. Samples were then dehydrated (70% IMS-denatured alcohol x2, 90% IMS x2, 100% IMS x3, xylene x3, paraffin wax x3) before being blocked in wax. Fifteen transverse sections, 6 mm thick and 60 mm apart from the next section, were cut from each dentine sample, mounted on poly-lysine slides, and heat fixed prior to staining using the Brown and Brenn method⁶.

Penetration of bacteria into dentine was visualized by light microscopy at x400 magnification. The central point of the root section was identified and five fields of view radiating out from this point were examined for each of the 15 root sections. The extent of invasion was initially expressed as the tubule invasion index (TI)¹⁶, where 1 to 20 tubules (per field) invaded scored 1; 21-50 tubules invaded scored 2; and >50 tubules invaded scored 3. These scores were then converted to Tubule Invasion Factor (TIF) that took depth of invasion of tubules into account⁵. The TIF was obtained by multiplying the TI by the invasion depth score: x1, where invasion depth was ≤50 mm; x2, where ≥5 tubules per field showed invasion depth >50 mm; and x3, where ≥5 tubules per field were invaded to depth of ≥100 mm, as previously described ⁵.

Identification of bacteria

Two root pieces from each appliance were rinsed in sterile distilled water and stored at -80°C. One of the specimens in each case was used to optimize the methods for DNA extraction and amplification. Once this had been achieved, the second sample from each subject was transferred into a microfuge tube containing 0.1 mL sterile 10% EDTA (pH 6.5), vigorously vortex-mixed, and incubated at room temperature for 30 min to partially decalcify. Samples were then transferred into 0.1 mL 2 M citric acid (pH 1.6) and incubated for 30 min. Samples were extensively rinsed in sterile distilled H₂O and transferred to tubes containing Gene Releaser (Cambio, Cambridge, Cambs, UK) for DNA extraction according to the manufacturer’s instructions.

The DNA extracts were used as templates in Polymerase Chain Reaction (PCR) amplification with universal 16S rRNA gene primers: DGGE F3(5’CGCCCGCCGCGCGCGGCGGGCGGGGCGGGGGCACGGGGGGCCTACGGGAGGCAGCAG) with the GC clamp for Denaturing Gradient Gel Electrophoresis (DGGE)²⁵ and R2 (5’ATTACCGCGGCTGCTGG), to amplify a product of 160 bp. The presence of correct sized fragments was confirmed by agarose gel electrophoresis. Subsequently, aliquots (6 mL) of PCR products were subjected to DGGE (50-60% denaturant gradient) and the separated bands were ethidium bromide-stained and visualized under UV light (344 nm) (Figure 2A). Bands were excised from the gel lanes, transferred to tubes containing 0.3 mL TE buffer (5 mM Tris-HCl, 10 mM EDTA, pH 7.5) and the DNA was allowed to elute from the gel fragments for 16 h at 4°C.

Figure 2
DGGE gel of DNA samples extracted from four (1-4) root dentine blocks (A) and agarose gel (B) showing PCR products derived from two selected DGGE gel bands from each sample (1-4). DNA bp markers (M) are indicated

Each of the eluted gel bands was then subjected to further PCR amplification using primers F3 (5’CCTACGGGAGGCAGCAG) (as above minus GC clamp), and R2 (above). Presence of amplified fragments was confirmed by agarose gel electrophoresis (Figure 2B). The fragments were gel-purified (QIAquick PCR Purification Kit, Qiagen, Manchester, Lancs, UK), ligated into plasmid pCR2.1 (Invitrogen, Thermo Fischer Scientific Inc., Waltham, MA, USA) and transformed into Escherichia coli XL1-Blue by standard procedures. Plasmids were extracted from transformant colonies using QIAprep Spin MiniPrep Kit (Qiagen), checked by agarose gel electrophoresis, and the 160-bp inserts were dideoxy-sequenced (Geneservice Ltd., Cambridge, Cambs, UK). The partial 16S rRNA gene sequences were then compared with 16S rRNA gene sequences in GenBank using the standard nucleotide NCBI/BLAST program.

RESULTS

Microscopic analysis of bacterial invasion

The tooth root dentine pieces were mounted into appliances as shown in Figure 1. The samples were designated A-D (lower right) (Figure 1B) and E-H (lower left). After 15 d in vivo, the root pieces were removed and processed as described in Material and Methods for microscopic analyses (A-F). All samples containing patent dentinal tubules showed high levels of bacterial invasion, with TIF values in the range of 5 to 9 (Figure 3). Three dentine samples (Figure 3) could not be assessed for invasion because of disintegration of internal dentine structure.

Figure 3
Mean level of microbial cell invasion into dentine samples carried by four subjects (1-4). Columns A through F correspond to dentine samples (see Figure 1 legend). Invasion is expressed as Tubule Invasion Factor (TIF) (see Material and Methods) which takes into account numbers of tubules containing bacteria and depth of penetration. Samples B, F, and B in subjects 2, 3, and 4 respectively, were not analyzable because of deformed dentine structures. Panel 5, combined dataset mean±standard deviation (8.09±0.87). Error bars are ± standard deviation from microscopic analysis of 75 individual sections (n=21)

The histochemical staining profiles with respect to type of organisms present within the dentinal tubules, pattern of tubule penetration, depth of invasion, and surface adhesion (biofilm) varied enormously across the dentine samples. We examined all dentine samples to confirm bacterial invasion, but in the following descriptions we have included only representative micrographs exhibiting distinct features of the invasion processes.

For subject 1, root sample D, there was invasion of purple-stained Gram-positive cocci (Figure 4A, arrowed a) and pink-stained Gram-negative rods (Figure 4A, arrowed b) to depths of >100 mm. Root sample F, on the other hand, seemed to be entirely permeated by small cocci bacteria. These stained Gram-positive in areas of denser colonization (Figure 4B, arrowed a) or Gram-negative in regions of deeper (~150 mm) invasion (Figure 4B, arrowed b). This Gram-variable staining was seen previously in laboratory studies of dentine invasion by pure cultures of streptococci¹⁶.

Figure 4
Transverse sections of human roots after 15 days incubation in situ in subject 1. Sections were prepared as described in Material and Methods, and stained by Brown & Brenn method. Panels: A, sample D, Gram-positive bacteria (a) and Gram-negative rods (b) penetrated to ~100 mm; B, sample F, small cocci infiltrated throughout the dentinal tubules appearing Gram-positive (a) towards the outside and Gram-negative (b) more internally. TIF scores for specimens are shown in Figure 3

In subject 2, sample A carried a dense invasive biofilm of Gram-positive stained material at the surface (Figure 5A) and there was invasion of tubules >150 mm by Gram-positive cocci. A similar pattern was seen for root sample D (Figure 5D), while sample C from subject 2 showed Gram-positive bacteria at the surface and deep penetration ≥200 mm by small Gram-negative organisms. Block B from this subject was one of the samples that could not be properly analyzed, as the internal dentine structure was disintegrated (Figure 5B).

Figure 5
Transverse sections of human roots after 15 days incubation in situ in subject 2. Sections were prepared as described in Material and Methods, and stained by Brown & Brenn method. Panels: A, sample A, Gram-positive cocci invading to a depth of >150 mm and showing a 10 mm depth dense biofilm at the surface (arrowed); B, sample B, disintegration of internal dentine structure meant that sections from this sample could not be analyzed; C, sample C, Gram-positive bacteria at the surface and deep penetration by smaller Gram-negative bacteria ≥200 mm (arrowed); sample D, Gram-positive and Gram-negative bacteria penetration with accumulation of Gram-positive cocci biofilm at the surface of the sectioned sample (arrowed). TIF scores for specimens are shown in Figure 3

Figure 6 shows sections from blocks C, D, and E from subject 3. Sections through C (Figure 6A) contained small Gram-negative rods within the dentinal tubules. A band of Gram-negative rods was present a small distance away from the surface of the dentine sample, perhaps having been present on the dentine surface prior to sectioning (Figure 6A). Sample D contained Gram-positive and Gram-negative rods (~5 mm length) in well-separated tubules and penetrating ≥150 mm (Figure 6B). In sample E, individual tubules contained deep lines of invading Gram-positive and Gram-negative bacteria (Figure 6C).

Figure 6
Transverse sections of human roots after 15 days incubation in situ in subject 3. Sections were prepared as described in Material and Methods, and stained by Brown & Brenn method. Panels: A, sample C, invasion by Gram-negative rod-shaped bacteria, with a strip of Gram-negative rods ~30 mm from the surface (arrowed); B, sample D, larger Gram-positive and Gram-negative rods (~5 mm length) well-separated but penetrating ≥150 mm; C, sample E, individual tubules appear to show long lines of invading Gram-positive and Gram-negative bacteria. TIF scores for specimens are shown in Figure 3

From subject 4, sample D showed invasion by Gram-positive cocci and Gram-negative rods, and a dense biofilm on the surface comprised of Gram-positive cocci and matrix material (Figure 7A). Sample E showed distinct penetration of tubules by groups of Gram-positive cocci (Figure 7B).

Figure 7
Transverse sections of human roots after 15 days incubation in situ in subject 4. Sections were prepared as described in Material and Methods, and stained by Brown & Brenn method. Panels: A, sample D, shows invasion by Gram-positive cocci and Gram-negative rods, together with a thick biofilm on the surface (arrowed) comprised of Gram-positive cocci and matrix material, staining pink; B, sample E, penetration by groups of Gram-positive cocci. TIF scores for specimens are shown in Figure 3

These results demonstrated that the dentine samples mounted onto the appliances were all readily susceptible to infection by invading microorganisms. The technique therefore was a very effective mean of achieving invasion of dentinal tubules by a variety of different oral bacteria.

Molecular identification of invading bacteria

Within one of the dentine samples selected at random from subject 1 (3 DGGE bands, Table 1) we identified principally Gram-negative bacteria including Klebsiella pneumoniae, Enterobacter species, Enterobacter hormaechei, and sequences similar to those from some uncultivated bacteria from faeces (Table 1). There was 100% sequence match over 160 bp to E. hormaechei. Subject 2 sample contained only Gram-negative Klebsiella-like bacteria, while subject 3 (three DGGE gel bands) provided sequences 100% identical to Acinetobacter and Streptococcus database entries, and 99% to Enterobacter spp. (Table 1). The root sample from subject 4 provided sequences with 100% matches to Gram-positive bacteria Granulictaella, Streptococcus mitis, Streptococcus oralis, and S. gordonii, and to Pseudomonas species and uncultivated organisms. Overall, these analyses showed a diversity of bacterial infection to a degree similar to morphological varieties of Gram-negative and Gram-positive organisms visualized microscopically (Figures 4-7).

Table 1
Microorganisms identified from within dentinal tubules following DNA extraction, PCR, DGGE, and 16S rDNA sequencing. The partial 16S rRNA sequences were compared using BLASTN with 16S rRNA gene sequences within GenBank

Subject	DGGE band	GenBank description1	No. 100% matches2,4	GenBank entry example3
1	U1	Klebsiella pneumoniae	0	KP297466
		Enterobacter hormaechei		KP027682
	U2	Enterobacter spp.	499	KP091277
		Enterobacter hormaechei		KF516241
	L2	Uncultivated from faeces	>500	KF841982
2	U1	Klebsiella oxytoca	499	CP004887
3	U1	Acinetobacter ursingii	69	LC014147
		Uncultured from skin		KF083053
	L1	Uncultured Streptococcus from skin/nasopharynx	248	KF505347
	L2	Enterobacter hormaechei	4	KF516241
4	U1	Granulicatella spp.	>500	KJ575555
	U2	Uncultured Pseudomonas spp.	7	AY191342
		Pseudomonas putida	0	KP114213
	L1	Streptococcus gordonii (from infective endocarditis)	>500	KJ170416
	L2	Streptococcus mitis	>500	KP233800
		Streptococcus oralis		LN589729
	L4	Uncultured human mouth	333	JQ457994
		Streptococcus sanguinis		AY944229

• ¹ Representative entries from the match listing
• ² Number of BLAST sequences with 100% match (160 bp)
• ³ GenBank Accesion numbers
• ⁴ 0 indicates 99% match (159/160)

DISCUSSION

In this study we have prepared dentine samples in a manner similar to that done for in vitro invasion investigations^16,22 of dentine infection by pure cultures of bacteria such as E. faecalis and Streptococcus species. These approaches have been undertaken to study the mechanisms involved in dentinal tubule infection, and to investigate the effects of various antiseptics, irrigants, and antimicrobials in preventing dentinal tubule infection. Perhaps one limitation of such in vitro analyses is that they have been undertaken under conditions that are quite different from those that would be encountered in vivo. These include, for example, the presence of whole saliva, salivary flow, shear and abrasion, and nutrient pulses. Our studies here show that it is possible to readily achieve dentine infection in vivo to the levels and extent that can be obtained in vitro¹⁶. This model therefore would be useful for testing the effects of new dentinal tubule occluding compounds²⁷ or agents for preventing root caries³⁰ in order to complement the in vitro experiments that have been previously employed.

Under laboratory conditions, Gram-positive cocci readily penetrate dentinal tubules. Historically, E. faecalis has been considered as a major invader of dentine^13,15, but more recent molecular studies that do not employ cultivation methods suggest that E. faecalis may not be so prevalent as generally believed²³. Invasions of dentine have been shown to contain a complex microbiota of Gram-positive and mainly Gram-negative bacteria⁸. Penetration of dentine by Gram-negative bacteria in vitro has not been investigated in such detail. Interestingly, periodontal bacteria Porphyromonas gingivalis were found to be unable to invade dentine unless co-cultured with Streptococcus¹⁶. In this present article we have demonstrated microscopically, and by molecular means, that dentine in vivo can be invaded by Gram-negative bacteria, principally Gram-negative rods. Some of the organisms identified, e.g., Enterobacter, Klebsiella, seemed on first impression to perhaps be unusual. However, Enterobacter and Klebsiella species have been identified within the subgingival microbiota⁸. More recently, Klebsiella was identified in deep carious lesions underneath restorations²⁰ and E. hormaechei was cultivated from human atherosclerotic tissue²⁴. Members of the Enterobacteriaceae and Pseudomonadaceae are also found on the human tongue⁹. Our work thus provides further evidence that these Gram-negative organisms are found in the oral cavity and have the ability to penetrate dentine.

A range of bacterial species were present within a small number of dentine samples analyzed. Three samples showed disintegration of tubule structure, most likely arising from the lengthy preparation process (fixation, demineralization, dehydration, sectioning). Only a limited number of specimens were employed here because we were interested in first establishing a model system. The results suggest that the model can be applied to future studies of dentine hypersensitivity agents, determining their clinical efficacy and their ability to occlude tubules and block bacterial invasion²⁷. It is acknowledged that the molecular methods used here do not differentiate between live or dead bacteria. However, it might be possible to utilize dentine discs, fracture them, and stain the intra-tubular bacteria with LIVE/DEAD stain. This method has recently been described in studies evaluating in vitro the antimicrobial effect of a commercial product on residual bacteria in dentinal tubules¹¹.

One of the samples in the study described here was invaded by several species of Gram-positive cocci, which corroborates the notion that these organisms are often some of the first to invade dentine¹⁵. However, E. faecalis was not found in our analyses. We identified Granulicatella, S. oralis, S. mitis, and S. gordonii which, with the exception of Granulicatella, have been previously implicated in tubule invasion¹⁵. In addition, all of these bacteria including Granulicatella are organisms that have been linked with infective endocarditis. Therefore, there could potentially be an association between ability to invade dentine and ability to cause endocardial or intravascular infections⁷.

We have thus identified organisms that were present within dentinal tubules that have been exposed to many hundreds of different bacteria in vivo¹⁹. In this study we only utilized four dentine samples to identify bacteria types that could invade the specimens under the condition used. Future clinical studies for testing efficacy of compounds or products in occluding tubules and preventing bacteria invasion would definitely employ many more subjects to provide suitable power. However, the molecular studies cannot be directly related to the morphological studies at this stage. We have established though that it is feasible to extract bacterial DNA from decalcified dentine. Our methodology would tend to identify the most prevalent microorganisms that were present within the dentine samples analyzed. We would like to develop these studies further in such a way that we could visualize and identify, by molecular techniques, the bacteria that have invaded the same dentine sample. This could be achieved by extracting bacterial DNA, or by detecting DNA using fluorescent in situ hybridization (FISH), from adjacent sections to those histochemically stained.

CONCLUSION

In summary, this study has established a novel in vivo model for studying the infection of dentine by oral microorganisms. Dentine specimens exposed to the human oral environment become infected with microorganisms to similar extent and depth to dentine infected in vitro under laboratory conditions. In addition to streptococci, bacteria from the genera Enterobacter, Klebsiella and Pseudomonas were identified as primary invading organisms. This in vivo model should provide the means to confirm in vitro experimental results on the effects of antiseptics, irrigants, or tubule occluding agents on dentine invasion by oral bacteria.

ACKNOWLEDGMENTS

We would like to thank Valeria Soro and Maria Davies for their excellent assistance. We are most grateful to the University of Bristol Dental School Technicians for the skilled preparation of clinical appliances. This study was financed in part by the University of Bristol Enterprise Development Fund.

REFERENCES

• ¹
  - Adriaens PA, De Boever JA, Loesche WJ. Bacterial invasion in root cementum and radicular dentin of periodontally diseased teeth in humans. A reservoir of periodontopathic bacteria. J Periodontol. 1988;59:222-30.
• ²
  - Arnold WH, Prange M, Naumova EA. Effectiveness of various toothpastes on dentine tubule occlusion. J Dent. 2015;43:440-9.
• ³
  - Arthur RA, Martins VB, Oliveira CL, Leitune VC, Collares FM, Magalhães AC, et al. Effect of over-the-counter fluoridated products regimens on root caries inhibition. Arch Oral Biol. 2015;60:1588-94.
• ⁴
  - Borges FM, Melo MA, Lima JP, Zanin IC, Rodrigues LK. Antimicrobial effect of chlorhexidine digluconate in dentin: in vitro and in situ study. J Conserv Dent. 2012;15:22-6.
• ⁵
  - Brittan JL, Sprague SV, Huntley SP, Bell CN, Jenkinson HF, Love RM. Collagen-like peptide sequences inhibit bacterial invasion of root dentine. Int Endod J. 2015; doi: 10.1111/iej.12474. Epub ahead of print.
• ⁶
  - Brown JH, Brenn L. A method for the differential staining of Gram-positive and Gram-negative bacteria in tissue sections. Bull Johns Hopkins Hosp. 1931;48:69-73.
• ⁷
  - Cargill JS, Scott KS, Gascoyne-Binzi D, Sandoe JA. Granulicatella infection: diagnosis and management. J Med Microbiol. 2012;61:755-61.
• ⁸
  - Chen L, Qin B, Du M, Zhong H, Xu Q, Li Y. et al. Extensive description and comparison of human-supragingival microbiome in root caries and health. PLoS One. 2015;10:e0117064.
• ⁹
  - Conti S, Santos SS, Koga-Ito CY, Jorge AO. Enterobacteriaceae and pseudomonadaceae on the dorsum of the human tongue. J Appl Oral Sci. 2009;17:375-80.
• ¹⁰
  - Fagrell TG, Lingström P, Olsson S, Steiniger F, Norén JG. Bacterial invasion of dentinal tubules beneath apparently intact but hypomineralized enamel in molar teeth with molar incisor hypomineralization. Int J Paediatr Dent. 2008;18:333-40.
• ¹¹
  - Hamama HH, Yiu CK, Burrow MF. Effect of silver diamine fluoride and potassium iodide on residual bacteria in dentinal tubules. Aus Dent J. 2015;60:80-7.
• ¹²
  - Hancock HH 3rd, Sigurdsson A, Trope M, Moiseiwitsch J. Bacteria isolated after unsuccessful endodontic treatment in a North American population. Oral Sur Oral Med Oral Pathol Oral Radiol Endod. 2001;91:579-86.
• ¹³
  - Love RM. Enterococcus faecalis - mechanism for its role in endodontic failure. Int Endod J. 2001;34:399-405.
• ¹⁴
  - Love RM, Chandler NP, Jenkinson HF. Penetration of smeared or nonsmeared dentine by Streptococcus gordonii Int Endod J.1996;29:2-12.
• ¹⁵
  - Love RM, Jenkinson HF. Invasion of dentinal tubules by oral bacteria. Crit Rev Oral Biol Med. 2002;13:171-83.
• ¹⁶
  - Love RM, McMillan MD, Park Y, Jenkinson HF. Coinvasion of dentinal tubules by Porphyromonas gingivalis and Streptococcus gordonii depends upon binding specificity of streptococcal antigen I/II adhesin. Infect Immun. 2000;68:1359-65.
• ¹⁷
  - Marcenes W, Kassebaum NJ, Bernabé E, Flaxman A, Naghavi M, Lopez A, et al. Global burden of oral conditions in 1990-2010: a systematic analysis. J Dent Res. 2013;92:592-7.
• ¹⁸
  - Nair PN, Sjögren U, Krey G, Kahnberg KE, Sundqvist G. Intraradicular bacteria and fungi in root-filled, asymptomatic human teeth with therapy-resistant periapical lesions: a long-term light and electron microscope follow-up study. J Endod. 1990;16:580-8.
• ¹⁹
  - Nasidze I, Li J, Quinque D, Tang K, Stoneking M. Global diversity in the human salivary microbiome. Genome Res. 2009;9:636-43.
• ²⁰
  - Neelakantan P, Rao CV, Indramohan J. Bacteriology of deep carious lesions underneath amalgam restorations with different pulp-capping materials – an in vivo analysis. J Appl Oral Sci. 2012;20:39-45.
• ²¹
  - Nyvad B, Fejerskov O. An ultrastructural study of bacterial invasion and tissue breakdown in human experimental root-surface caries. J Dent Res. 1990;69:1118-25.
• ²²
  - Ørstavik D, Haapasalo M. Disinfection by endodontic irrigants and dressings of experimentally infected dentinal tubules. Endod Dent Traumatol. 1990;6:142-9.
• ²³
  - Ozok AR, Persoon IF, Huse SM, Keijser BJ, Wesselink PR, Crielaard W, et al. Ecology of the microbiome of the infected root canal system: a comparison between apical and coronal root segments. Int Endod J. 2012;45:530-41.
• ²⁴
  - Rafferty B, Dolgilevich S, Kalachikov S, Morozova I, Ju J, Whittier S, et al. Cultivation of Enterobacter hormaechei from human atherosclerotic tissue. J Atherscler Thromb. 2011;18:72-81.
• ²⁵
  - Sheffield VC, Cox DR, Lerman LS, Myers RM. Attachment of a 40-base-pair G + C-rich sequence (GC-clamp) to genomic DNA fragments by the polymerase chain reaction results in improved detection of single-base changes. Proc Nat Acad Sci U S A. 1989;86:232-6.
• ²⁶
  - Siqueira JF Jr, Rôças IN. Diversity of endodontic microbiota revisited. J Dent Res. 2009;88:969-81.
• ²⁷
  - Wang R, Wang Q, Wang X, Tian L, Liu H, Zhao M, et al. Enhancement of nano-hydroxyapatite bonding to dentin through a collagen/calcium dual-affinitive peptide for dentinal tubule occlusion. J Biomat Appl. 2014;29:268-77.
• ²⁸
  - West NX, MacDonald EL, Jones SB, Claydon NC, Hughes N, Jeffery P. Randomized in situ clinical study comparing the ability of two new desensitizing toothpastes technologies to occlude patent dentin tubules. J Clin Dent. 2011;22:82-9.
• ²⁹
  - Xhevdet A, Stubljar D, Kriznar I, Jukic T, Skvarc M, Veranic P, et al. The disinfecting efficacy of root canals with laser photodynamic therapy. J Lasers Med Sci. 2014;5:19-26.
• ³⁰
  - Zhang N, Melo MA, Chen C, Liu J, Weir MD, Bai Y, et al. Development of a multifunctional adhesive system for prevention of root caries and secondary caries. Dent Mater. 2015;31:1119-31.

Publication Dates

• Publication in this collection
  Mar-Apr 2016

History

• Received
  9 Oct 2015
• Reviewed
  10 Jan 2016
• Accepted
  21 Jan 2016