
Water output from dental unit waterlines (DUWLs) is used for various dental treatments. DUWL water should always be kept clean because it comes into direct or indirect contact with patients and dental workers during dental treatments. The American Dental Association recommends that microbial contamination levels in DUWL water should be kept below 200 colony forming units (CFU) per 1 ml1). However, previous studies have shown that microbial contamination in DUWL water exceeds the recommended level (200 CFU/ml)2-4). The primary cause of microbial contamination in DUWL water is biofilm formation on the inner surface of DUWLs5); therefore, these biofilms must be removed to reduce microbial contamination in DUWLs and provide safer dental treatments for patients. While various methods are used to reduce microbial contamination in DUWLs to the recommended level, methods that do not directly mitigate DUWL biofilms only temporarily lower microbial contamination levels, causing the microbial contamination levels to increase again within a short period6-8).
A biofilm is an assemblage of microorganisms irreversibly bound to a surface and enclosed in a matrix composed mainly of polysaccharide material9). The attachment of bacteria to a surface is the first step in the formation of a biofilm10). Early colonizing bacteria attached to surfaces are very important as they provide attachment substrates for later-colonizing bacteria and ultimately influence subsequent biofilm formation11). This means that a simple way to prevent biofilm formation is to inhibit the attachment of bacteria to surfaces12). Therefore, information on the species and characteristics of early colonizing bacteria that initiate biofilm formation is necessary to inhibit biofilm formation and maturation.
In dental plaque, a well-known biofilm, culture-based techniques have revealed that streptococci, and Neisseria and Rothia species dominate early colonizing bacteria13). However, there are limitations in fully identifying the bacteria that compose a biofilm through culture-based techniques; therefore, efforts are ongoing to identify early colonizing bacteria through techniques based on bacterial DNA11,13,14). Diaz et al.14) placed enamel chips on healthy human subjects and performed an RNA-based analysis of bacterial species from dental plaque formed after 4 and 8 hours. Unculturable species were found for the first time in this study.
Similarly, the bacteria that comprise DUWLs biofilms have also been studied. Most studies have investigated bacterial diversity within DUWLs biofilms but have not identified early colonizing bacteria. In addition, culture- and DNA-based techniques have been used to identify bacterial diversity in DUWLs biofilms15-18). However, the differences in the results revealed by each technique indicate that further research is required to identify the bacteria in DUWLs biofilms.
The thymine adenine (TA) cloning technique was used to clone the 16S ribosomal DNA (rDNA) of a bacterium, build a library, and sequence it to identify the bacterium19).
This study aimed to identify the early colonizing bacteria in the DUWLs biofilm formation process by identifying the bacterial species formed in the DUWLs biofilm model according to the formation period using TA cloning.
DUWL biofilms were formed in the United States Centers for Disease Control and Prevention (CDC) biofilm reactor model established in a previous study20). The amount of biofilm accumulation after various biofilm formation periods was identified, and based on this amount, biofilm samples at 5, 45, and 150 minutes and 1 and 4 days of biofilm formation were used for TA cloning analysis.
Biofilm samples formed on polyurethane tubing were suspended in 1 ml of phosphate-buffered saline (PBS). The suspension was centrifuged for 1 minutes at 13,000×g, and the samples were resuspended in 200 μl of PBS. Genomic DNA was extracted from the resuspension solution using a G-spin Genomic DNA Extraction Kit (iNtRON Biotechnology, Seongnam, Korea) following the manufacturer’s protocol. Universal polymerase chain reaction (PCR) primer (27F; 5’-AGA GTT TGA TCM TGG CTC AG-3’, 1492R; 5’-GGY TAC CTT GTT ACG ACT T-3’) and 2X Taq PCR Premix (SolGent Co. Ltd., Daejeon, Korea) were used to amplify 16S rRNA genes in GeneAmp PCR System 9700 (PerkinElmer, Waltham, MA, USA). PCR was performed under the following conditions: after adding 0.1 µM of forward and reverse primers and 100 pg of bacterial genomic DNA, distilled water was added until a final concentration of 30 µl was achieved. The sample was treated for 2 minutes at 94°C, followed by 30 seconds of denaturation at 94°C, 30 seconds of annealing at 55°C, and 1 minute of extension at 72°C (34 cycles). The final product was electrophoresed on a 1% agarose gel and purified using an AccuPrep Gel Purification Kit (BIONEER, Daejeon, Korea) following the manufacturer’s protocol. Gel-purified 16S rDNA was cloned using MG TA TOPO Cloning Kit (MGmed-Doctor Protein, Seoul, Korea) following the manufacturer’s protocol, which was followed by transformation using MG DH5α competent cells (MGmed-Doctor Protein). Escherichia coli cells containing the cloned recombinant plasmids were plated on Luria-Bertani (LB) agar plates (Becton Dickinson & Co., Sparks, MD, USA) containing ampicillin and incubated, after which 30∼40 colonies per biofilm formation period were randomly selected.
After incubating the selected colonies in 3 ml of LB broth, plasmids were extracted using the AccuPrep Plasmid Nano-Plus Plasmid Mini Extraction Kit (BIONEER) following the manufacturer’s protocol. Extracted plasmid DNA was treated with the restriction enzyme EcoR1 (Invitrogen, Carlsbad, CA, USA), and inserted DNA was checked using electrophoresis, after which Cosmogenetech (Seoul, Korea) was contracted for sequencing. The nucleic acid sequence of the analyzed 16S rDNA was used for a homology search in the GenBank database (http://www.ezbiocloud.net). Among the species with ≥98% homology, species with the highest rate of homology were determined as target species.
Fig. 1 shows the time-dependent accumulation of biofilms in the CDC biofilm reactor. The amount of biofilm accumulation increased at 5, 45, and 150 minutes and 1 and 4 days of biofilm formation.
TA cloning analysis was performed on the batch culture broth before biofilm formation and on samples of biofilms formed over 5, 45, and 150 minutes, and 1 and 4 days. Among 267 colonies, 13 genera and 19 species were identified (Table 1). For Hydrotalea flava, the highest number of clones was isolated in batch culture. Rhizobium pusense and Pseudomonas parafulva had fewer clones isolated from batch cultures but an increase in the number of clones isolated from biofilms. Halomonas hamiltonii had the highest number of clones at 5 minutes after biofilm formation, but fewer clones were isolated from biofilms grown for 45 minutes, 150 minutes, 1 day, or 4 days. For P. parafulva, the count increased up to 150 minutes of biofilm formation, followed by a decrease. For Bradyrhizobium spp., the count increased on 1 day of biofilm formation.
Summary of Isolated Clones Derived from Dental Unit Waterlines Biofilms
Phylum | Batch culture | Biofilm formation | |||||
---|---|---|---|---|---|---|---|
Genus | Species | 5 min | 45 min | 150 min | 1 d | 4 d | |
Actinobacteria | |||||||
Leifsonia | Leifsonia shinshuensis | 0 | 1 | 0 | 0 | 0 | 0 |
Microbacterium | Microbacterium oxydans | 0 | 0 | 0 | 0 | 0 | 1 |
Bacteroidetes | |||||||
Hydrotalea | Hydrotalea flava | 68 | 0 | 1 | 0 | 0 | 0 |
Proteobacteria | |||||||
Acidovorax | Acidovorax soli | 1 | 0 | 0 | 0 | 0 | 0 |
Afipia | Afipia broomeae | 4 | 0 | 0 | 0 | 0 | 0 |
Afipia massiliensis | 1 | 0 | 0 | 0 | 0 | 0 | |
Bradyrhizobium | Bradyrhizobium spp. | 4 | 2 | 2 | 2 | 6 | 4 |
Brevundimonas | Brevundimonas spp. | 0 | 0 | 0 | 1 | 0 | 2 |
Halomonas | Halomonas hamiltonii | 4 | 10 | 2 | 8 | 3 | 3 |
Halomonas stevensii | 0 | 0 | 3 | 4 | 0 | 1 | |
Methylobacterium | Methylobacterium fujisawaense | 0 | 0 | 0 | 0 | 0 | 1 |
Methylobacterium oryzae | 0 | 0 | 0 | 0 | 1 | 1 | |
Polaromonas | Polaromonas aquatica | 1 | 0 | 0 | 0 | 0 | 0 |
Pseudomonas | Pseudomonas fulva | 5 | 5 | 5 | 2 | 5 | 3 |
Pseudomonas parafulva | 2 | 6 | 8 | 9 | 7 | 7 | |
Pseudomonas vranovensis | 1 | 0 | 0 | 0 | 0 | 0 | |
Rhizobium | Rhizobium larrymoorei | 0 | 0 | 0 | 0 | 2 | 0 |
Rhizobium pusense | 2 | 8 | 11 | 9 | 8 | 6 | |
Shigella | Shigella spp. | 0 | 0 | 0 | 0 | 0 | 1 |
No hits found | 4 | 1 | 2 | 1 | 3 | 3 | |
Total | 97 | 33 | 34 | 36 | 35 | 33 |
Values are presented as number only.
DUWL biofilms comprise various bacterial species. Using the latest high-throughput DNA sequencing technology, 63∼394 bacterial genera have been identified17,18,21). Despite many studies on identifying bacteria in DUWLs using various techniques, none have identified the constituent bacterial strains in biofilms during DUWL biofilm formation. Bacterial adhesion to DUWLs is the first and most important step in the DUWL biofilm formation process10,22). To identify the bacteria involved in the early adhesion step of the DUWL biofilm formation process, we identified the constituent bacterial strains at different biofilm formation periods. This study is the first to identify the constituent bacterial strains in DUWL biofilms.
High-throughput DNA sequencing, which has been used in recent years to identify bacterial diversity, offers the efficiency of obtaining many bacterial DNA sequences within a short time; however, it is difficult to identify these strains accurately at the species level because the average size of the bacterial base sequences is approximately 350 bp17). Because of these disadvantages, this study used the TA cloning technique to identify bacterial strains. TA cloning is a culture-independent molecular approach in which the 16S rDNA of a bacterium is cloned23). TA cloning has the advantage of identifying bacteria that are not cultured and even quantitatively identifying bacteria from the sample24). In addition, TA cloning provides highly accurate identification because it is possible to obtain bacterial sequences of sufficient length24). In our study, using TA cloning, we identified bacteria comprising the initial DUWLs biofilm at the species level, including uncultured bacteria, and quantitatively determined the presence of bacteria during biofilm formation.
In this study, early adherent bacteria were identified by the formation of DUWLs biofilms in a laboratory using a CDC reactor. As laboratory biofilm models provide a similar and repeatable environment for biofilm formation, they can be utilized in studies that specifically characterize biofilms and determine the effectiveness of different disinfectants25,26). The CDC for Disease Control and Prevention biofilm reactor, developed at the Biofilm Engineering Center at Montana State University, was used to generate biofilms in the laboratory. Because the magnetic stir bar in the center of the CDC Biofilm Reactor allows the replication of dynamic conditions during biofilm formation, it may be better suited to replicate biofilms formed under dynamic conditions such as saliva or water flow20). Since DUWLs biofilms also form in the presence of water flow, the CDC biofilm reactor model developed by the authors was similar to clinical DUWLs biofilms and utilized to simulate DUWLs biofilms20). Using a laboratory model, it may be possible to isolate and obtain quantitative data on early adherent bacteria from DUWLs biofilms formed in similar environments. When reproducing biofilms in this laboratory model and measuring the amount of biofilm accumulation, the 4-day biofilm showed the highest accumulation. Because most dental offices are supplied with tap water low in bacterial contamination and nutrients, 4-day old dental DUWLs do not exhibit the biofilm accumulation observed in this model25,27). The laboratory model was based on a batch culture of bacteria fed with an R2A medium to reproduce a mature biofilm in a short period of time, which was different from the situation of DUWLs used in dental clinics for four days. Based on the DUWLs biofilms formed in this laboratory model, the biofilm formed by day 4 can be considered the initial biofilm, and biofilms from this period were targeted for identifying early adherent bacteria. Consequently, 267 clones, 13 genera, and 19 species were identified in the initially formed biofilm.
In our study, H. flava was dominant in batch cultures used for biofilm formation. However, it was barely detectable in biofilms formed in laboratory models. Of the approximately 19,000 genera detected in water discharged from DUWLs, Hydrotalea (0.5%) was detected, although not dominant18). A batch culture is a solution of bacteria collected from the water discharged from DUWLs. Therefore, Hydrotalea are thought to be isolated and derived from mature biofilms due to water flow within DUWLs. For this reason, H. flava can be speculated to play a role as a late colonizer in DUWLs biofilm formation. In contrast, in the same previous study, Rhizobium was one of the genera barely detected in the water discharged from DUWLs and was also detected in low numbers in the batch culture in our study18). However, biofilms formed over a short period had the highest detection rate. This suggests that R. pusense may act as an initial colonizer in forming biofilms in DUWLs. Similar to R. pusense, the detection of P. parafulva was low in batch culture but increased within the biofilm formed. It can be speculated that P. parafulva may also play the role of an early colonizer.
The genus Halomonas is dominant in water discharged from DUWLs17,18). In our study, biofilms that formed for 5 minutes had high detection rates; however, the number of detections decreased with longer biofilm formation periods. Li et al.11) showed that the bacterial composition of dental plaque immediately after pumicing changes over time and that the bacterial composition of dental plaque immediately after pumicing is similar to that of saliva. This was attributed to the loose attachment of bacteria in saliva due to exposure to saliva immediately after pumicing, which ultimately showed that only specific bacteria could adhere initially11). Similarly, in the DUWLs laboratory model, owing to media flow, Halomonas was loosely attached at the beginning of biofilm formation and dislodged from the media flow, resulting in reduced detection within the subsequently formed biofilms. H. hamiltonii does not produce extracellular polysaccharides28). Therefore, Halomonas is unlikely to be an early adherent bacterium.
Because bacterial adhesion is also influenced by properties such as hydrophobicity and extracellular polysaccharides, further research is needed on the properties that affect adhesion in these bacteria, especially R. pusense and P. parafulva, which are considered early adherent bacteria22,29,30).
To isolate the bacteria involved in early adhesion, suspensions of the biofilms formed at 5 and 45 minutes in PBS solution were cultured on R2A solid medium, and the colonies produced after incubation were pure cultured for genomic DNA extraction and sequencing; however, we did not find R. pusense and H. hamiltonii in any of the colonies produced (results not shown). The R2A medium is recommended for isolating bacteria from DUWLs31). R2A medium is used to isolate bacteria that live in low-nutrient environments such as water32). However, these results show that some bacteria in DUWLs cannot be cultured in R2A medium, suggesting the need for a suitable method for culturing bacteria living in DUWLs.
In previous studies that utilized culture-based techniques to identify bacterial diversity in DUWLs, Sphingomonas paucimobilis, Acinetobacter calcoaceticus, and Methylobacterium mesophilicum were the dominant bacteria4,33). In contrast, studies utilizing high-throughput DNA sequencing techniques to determine bacterial diversity in DUWLs revealed that Halomonas was the dominant genus, followed by Sphingomonas17,18). The reason why studies utilizing culture-based techniques have not found Halomonas to be dominant is also speculated to be due to the media used in each study. If the difficulties in identifying various bacteria involved in adhesion at the species level and the laboratory culture of adhering bacteria are resolved, it would be possible to develop methods to identify the underlying adhesion mechanisms and prevent DUWL biofilm formation.
The discovery and application of DUWLs surface materials or disinfectants that inhibit the adhesion of early adherent bacteria identified in this study could inhibit the formation of DUWLs biofilms.
This study is the first to use TA cloning to identify bacterial strains in DUWL biofilms during formation. However, in TA cloning approaches, several factors can bias the inferred community structure, including total DNA extraction from the sample, PCR amplification bias, gene copy number bias, and cloning and sequencing artifacts23). This corresponds to each step of the TA cloning approach, and each step requires the implementation of a rigorous protocol by the researcher to reduce bias23). For this reason, there are some limitations to increasing the number of clones. However, in this study, the number of clones analyzed using TA cloning was somewhat low at 267, which is relatively low compared to the approximately 19,000 bacterial strains detected in the output water of DUWLs using a high-throughput DNA sequencing technique in our previous study. Therefore, future studies should be conducted using a larger number of clones for TA cloning analysis.
In addition, biofilms formed from laboratory models are based on inoculated bacterial cultures and are therefore limited in their ability to reproduce the DUWLs environment with a wide variety of microorganisms. Future studies should use high-throughput DNA sequencing techniques to identify the genera of the composing bacteria during each period of biofilm formation.
None.
Conflict of interest
No potential conflict of interest relevant to this article was reported.
Ethical approval
This article is not necessary for IRB screening because it is an experimental paper utilizing a laboratory model of biofilms based on bacteria collected at DUWL.
Author contributions
Conceptualization: Hye Young Yoon and Si Young Lee. Data acquisition: Hye Young Yoon and Si Young Lee. Formal analysis: Hye Young Yoon and Si Young Lee. Supervision: Hye Young Yoon. Writing-original draft: Si Young Lee. Writing-review & editing: Hye Young Yoon and Si Young Lee.
Funding
None.
Data availability
Please contact the corresponding author for data availability.
![]() |
![]() |