Filifactor alocis - involvement in periodontal biofilms
© Schlafer et al. 2010
Received: 21 August 2009
Accepted: 1 March 2010
Published: 1 March 2010
Skip to main content
© Schlafer et al. 2010
Received: 21 August 2009
Accepted: 1 March 2010
Published: 1 March 2010
Bacteria in periodontal pockets develop complex sessile communities that attach to the tooth surface. These highly dynamic microfloral environments challenge both clinicians and researchers alike. The exploration of structural organisation and bacterial interactions within these biofilms is critically important for a thorough understanding of periodontal disease. In recent years, Filifactor alocis, a fastidious, Gram-positive, obligately anaerobic rod was repeatedly identified in periodontal lesions using DNA-based methods. It has been suggested to be a marker for periodontal deterioration. The present study investigated the epidemiology of F. alocis in periodontal pockets and analysed the spatial arrangement and architectural role of the organism in in vivo grown subgingival biofilms.
A species-specific oligonucleotide probe, FIAL, was designed and evaluated. A total of 490 subgingival plaque samples were submitted to PCR and subsequent dot blot hybridization to compare the prevalence of F. alocis in patients suffering from generalized aggressive periodontitis (GAP), chronic periodontitis (CP), and control subjects resistant to periodontitis. Moreover, a specially designed carrier system was used to collect in vivo grown subgingival biofilms from GAP patients. Subsequent topographic analysis was performed using fluorescence in situ hybridization.
While the majority of patients suffering from GAP or CP harboured F. alocis, it was rarely detected in the control group. In the examined carrier-borne biofilms the organism predominantly colonized apical parts of the pocket in close proximity to the soft tissues and was involved in numerous structures that constitute characteristic architectural features of subgingival periodontal biofilms.
F. alocis is likely to make a relevant contribution to the pathogenetic structure of biofilms accounting for periodontal inflammation and can be considered an excellent marker organism for periodontal disease.
Periodontitis is a chronic inflammatory bacterial infection leading to destruction of periodontal ligaments and supporting bone of the tooth. Its aetiology has been a field of intensive research in the past decades. As periodontal pockets accommodate a multitude of bacterial phylotypes, it is difficult to differentiate between mere commensals and true pathogens. During the 1970's, 80's and early 90's, research focused mainly on a number of culturable bacteria like Porphyromonas gingivalis, Prevotella intermedia, Aggregatibacter (Actinobacillus)actinomycetemcomitans, Tannerella forsythia and Treponema denticola that proved to be associated with the disease . Studies have determined their relative prevalences, interactions and virulence factors [2–7]. By the end of the 1980's, the development of novel, culture-independent techniques allowed the identification of as-yet-unculturable and fastidious organisms in patients suffering from periodontitis and added new insight into bacterial communities in periodontal pockets [8–10]. In recent years, research has detected increasing numbers of bacterial species and phylotypes in subgingival plaque and other habitats of the human oral cavity [11–18]. There is little reason to believe that easily culturable bacteria contribute more to the development of periodontitis than fastidious organisms. Doubt has been raised whether the widely accepted periodontal pathogens P. gingivalis, P. intermedia and T. forsythia are appropriate diagnostic markers to differentiate between health and disease [19, 20].
Along with these discoveries it became clear that the mere isolation and characterization of bacteria from diseased sites is not a sufficient approach to understand the complex pathogenesis of periodontitis. The organisms do not live in a planktonic form, but rather as a sessile community attached to the tooth surface in a matrix of extracellular polymers . The structure and function of these bacterial biofilms are influenced both by bacterial interactions and host factors. Exploring the biofilm architecture and identifying its bacterial architects are pressing goals in current periodontal research.
Filifactor alocis (ATCC 35896T) was first isolated in 1985 from the human gingival crevice as Fusobacterium alocis  and later reclassified as Filifactor alocis . It is a fastidious, Gram-positive, obligately anaerobic rod that possesses trypsin-like enzymatic activity , as do P. gingivalis and T. denticola [25, 26]. In recent years, it has been discovered in patients suffering from chronic periodontitis (CP) [14, 18, 27, 28], generalized aggressive periodontitis (GAP)  and endodontic infections . Recently, F. alocis was detected in elevated numbers in CP patients with periodontal deterioration compared to patients with a stable periodontal condition and was therefore proposed as a potential marker for active disease .
The present study chose a DNA-based epidemiological approach utilizing dot blot hybridization to investigate the prevalence of F. alocis in subjects with GAP, CP, and in a subject group resistant to periodontitis. Furthermore, fluorescence in situ hybridization (FISH) was employed to analyse the spatial arrangement and the architectural role of F. alocis in periodontal pockets. For that purpose, a specially designed carrier system was used to collect in vivo grown biofilms from GAP patients .
To detect F. alocis, a species-specific probe, FIAL (5'-TCTTTGTCCACTATCGTTTTGA-3') was designed after comparative sequence analysis of close phylogenetic neighbours to F. alocis. To ensure specificity, the probe sequence was compared to the sequences deposited in the Ribosomal Database Project II  and to all 16S rRNA entries at the EMBL and GenBank databases (as of August 2009) employing the Husar program package (DKFZ, Heidelberg, Germany). The probe was checked for its practical use in hybridization experiments with the program OLIGO (version 4.0). EUB 338, a probe complementary to a highly conserved region of the 16S rRNA gene in bacteria, was used in dot blot hybridization experiments to verify successful PCR amplification and in FISH experiments to detect and visualize large parts of the bacterial biofilm population . For comparative purposes, probes POGI, PRIN, ACAC, TDEN, FUNU and B(T)AFO were employed in dot blot experiments to detect P. gingivalis, P. intermedia, A. actinomycetemcomitans, T. denticola, Fusobacterium nucleatum and T. forsythia, respectively. These probes have been published previously and deposited in ProbeBase .
Clinical criteria for patient selection
Periodontitis Resistant (PR) subjects
Age ≥ 65 years
≥ 20 natural teeth
Probing Depth at any site ≤ 5 mm
Clinical Attachment Loss at any site ≤ 2 mm
Chronic Periodontitis (CP)
≥ 4 mm Probing Depth at ≥ 30% of residual teeth
Generalized Aggressive Periodontitis (GAP)
Disease onset estimated at < 30 years based on clinical examination, past radiographs, and/or interview
≥ 6 mm Probing Pocket Depth at > 3 permanent teeth other than first molars and incisors
Clinical samples processed by dot blot hybridization
(mm) ± SD
34.8 ± 6.4
7.8 ± 2.5
51.0 ± 10.2
7.1 ± 1.4
66.7 ± 1.5
3.6 ± 0.8
Clinical samples for FISH
(mm) ± SD
34.3 ± 7.9
8.1 ± 1.7
Statistical evaluation of the dot blot hybridization results was performed using the exact chi-square test. The prevalence of F. alocis in different patient groups was compared. Moreover, the presence of F. alocis in relation to the PPD was analysed. P values below 0.05 were considered statistically significant.
A carrier system designed to collect biofilms grown in vivo in periodontal pockets was used for sampling . Ethics approval for subgingival sample collection was given by the Ethical Committee at Charité - Universitätsmedizin Berlin. Expanded polytetrafluoroethylene (ePTFE) membranes were placed in periodontal pockets of GAP patients for 7 to 14 days and colonized by the subgingival bacterial flora. Strips of ePTFE measuring 3 mm in width were wrapped around and attached to rigid plastic tips (Plast-O-Probe; Maillefer, Ballaigues, Switzerland), which permitted the insertion of these strips down to the bottom of the pocket, therefore allowing the complete extension of the membrane over the entire probing depth. One side of the double bent strip faced the soft tissue and the other side, slightly longer, faced the root surface. This longer cervical end was fixed to the tooth with cyanoacrylic glue (Tesa, Beiersdorf, Hamburg, Germany) to stabilize the position of the carrier. After removal, carriers were fixed for at least 3 h with 3.7% (v/v) formaldehyde in phosphate-buffered saline (pH 7.4) and embedded in cold polymerizing resin (Technovit 8100, Kulzer, Wehrheim, Germany) as reported previously . Sectioning into slices of 2-3 μm was performed as previously published . A total of 28 carriers from 11 GAP patients seeking treatment at the Charité - Universitätsmedizin Berlin were examined. These patients met the same inclusion criteria as the GAP patients selected for dot blot hybridization and likewise signed informed consent forms. See Table 2 for patient demographics. Additionally, a gingival biopsy of a GAP patient obtained during periodontal surgery was processed in the same manner and included in the FISH experiments.
FISH experiments were performed as described previously  apart from using Vectashield containing DAPI (4,6-Diamidino-2-Phenylindoldihydrochlorid) (Vector Laboratories, Orton Southgate, UK) as mounting medium. The probes were synthesized commercially (biomers.net, Ulm, Germany). EUB 338 was 5' end-labelled with fluorochrome Cy5 (indodicarbocyanine) while FIAL was 5' end-labelled with fluorochrome Cy3 (indocarbocyanine). Differential labelling allowed simultaneous hybridization with both probes.
The stringency of FIAL was adjusted by incubating fixed cells of F. alocis and its closest cultured relative, F. villosus with different hybridization mixes. The formamide concentrations covered a range from 0% (v/v) to 75% (v/v), rising in steps of 5% (v/v). At each level of formamide, a series of images of each bacterial species was taken with a fixed exposure time. The software daime  was used to measure the light intensities emitted by both species for each concentration of formamide. While the signal intensity of F. villosus did not reach 50 Relative fluorescence Units (RU) at any level of formamide due to unspecific binding of the probe, the intensity of F. alocis remained constantly above 150 RU using formamide concentrations of up to 20% (v/v) (see Additional file 1). In addition, fixed cells of 16 different bacterial species, most of them periodontal pathogens, were incubated with FIAL at 20% (v/v) formamide as negative controls, namely F. nucleatum (ATCC 25586), Eikenella corrodens (CCUG 2138), Kingella kingae (ATCC 23330), Veillonella parvula (ATCC 10790), Veillonella dispar (ATCC 17748), P. gingivalis (ATCC 33277), A. actinomycetemcomitans (ATCC 33384), Pasteurella haemolytica (ATCC 33396), T. forsythia (ATCC 43037), Haemophilus aphrophilus (NCTC 55906) P. intermedia (ATCC 25611), Campylobacter rectus (ATCC 33238), Capnocytophaga sputigena (ATCC 33612), Capnocytophaga gingivalis (ATCC 33624), Eggerthella lenta (ATCC 25559), and Peptostreptococcus anaerobius (ATCC 27337). As none of the controls were detected by FIAL, all further experiments were performed with 20% (v/v) of formamide, including F. alocis as positive and F. villosus as negative control.
After hybridization, carrier and biopsy sections were analysed using an epifluorescence microscope (AxioPlan II, Zeiss, Jena, Germany) equipped with a 100 W high pressure mercury lamp (HBO 103W/2, Osram, Munich, Germany) and 10×, 40× and 100× objectives. DAPI, Cy3 and Cy5 signals were analysed by narrow band filter sets HQ F31-000, HQ F41-007 and HQ F41-008, respectively (AHF Analysentechnik, Tübingen, Germany). Image acquisition was performed with an AxioCam MRm (Zeiss) making use of the AxioVision 4.4 software.
When carried out with the probe EUB 338 (specific for most bacteria), dot blot hybridization experiments indicated the presence of bacteria in all 490 patient samples as well as in the positive (F. alocis) and negative controls (see Figure 1 legend) and thus confirmed successful PCR amplification (Figure 1a). The Filifactor alocis-specific probe FIAL clearly detected F. alocis, while neither the closest phylogenetic neighbour F. villosus nor any of the organisms in the panel of oral bacteria (see Figure 1 legend) yielded a signal, thus indicating specific hybridization conditions (Figure 1b).
The signal intensity of the FIAL-positive patient samples varied between the three groups, suggesting a higher number of Filifactor in GAP and CP pockets than in PR pockets tested positive for the organism. Nonetheless, as hybridizations were carried out on PCR-amplified bacterial DNA, no further analysis of signal intensities was performed.
Detection frequencies of P. gingivalis, P. intermedia, A. actinomycetemcomitans, T. denticola, T. forsythia, and F. nucleatum in the three patient groups are displayed in Figure 2b.
To investigate the prevalence of F. alocis in relation to the PPD, the donor sites were divided into four groups (I: 1-3 mm, II: 4-6 mm, III: 7-9 mm, IV: > 9 mm). As there is a certain degree of interdependency between pockets belonging to the same patient, statistical analysis was limited to one pocket per patient and probing depth group. Although a slightly higher percentage of group III pockets than group II pockets was positive for Filifactor in both the GAP and the CP patients, these differences were not statistically significant. Similarly, analysis revealed no statistically significant differences in the prevalence of the organism in GAP patients compared to CP patients in both pockets of 4-6 mm and pockets of 7-9 mm. In contrast, the prevalence of F. alocis in pockets of 4-6 mm differed significantly between both PR and GAP patients (p < 0.001) and PR and CP patients (p < 0.001) (Figure 2c). Insufficient numbers or complete absence of pockets of 1-3 mm in GAP and CP patients, pockets of 7-9 mm in PR patients and pockets deeper than 9 mm in CP and PR patients did not permit further statistical analysis.
To our knowledge, the present study is the first to analyse the prevalence of F. alocis in samples from both GAP and CP patients, and subjects with apparent periodontitis resistance. The detection of the organism in 77.8% of the GAP patients and in 76.7% of those suffering from CP is convincing evidence that suggests an involvement of F. alocis in periodontal disease. Equally striking is the low prevalence of Filifactor in the PR group. All of these patients had reached the age of 65 years and were in good periodontal condition without the help of extensive therapeutic efforts. Even if a multitude of factors including oral hygiene and immune response contributed to their periodontal status, one would assume that frequent detection of an organism in the GAP and CP groups along with scarce detection in PR patients, as is the case for F. alocis, indicates pathogenic rather than commensal behaviour.
One can argue that deep periodontal pockets harbour increased numbers of bacteria and that any organism inevitably should be isolated more constantly from CP patients (mean pocket depth: 7.13 mm, 1.4 mm SD) and especially GAP patients (7.81 mm, 2.48 mm SD) than from PR patients (3.63 mm, 0.79 mm SD). However, dividing the entirety of the sampled sites into four groups according to the measured PPD (I: 1-3 mm, II: 4-6 mm, III: 7-9 mm, IV: > 9 mm), statistical analysis of pockets between 4 and 6 mm still reveals a significantly higher prevalence of F. alocis in both the GAP and the CP group compared to the PR group. In addition, the organism was not detected significantly more frequently in deeper pockets (7-9 mm) than in rather shallow pockets (4-6 mm) in both GAP and CP patients. Although a connection between PPD and bacterial load cannot be denied, these findings indicate that the influence of pocket depth does not invalidate the aforementioned results.
If one compares the prevalence rate of F. alocis to those of the widely accepted periodontal pathogens P. gingivalis, P. intermedia, A. actinomycetemcomitans, T. denticola, F. nucleatum, and T. forsythia (see Figure 2b), investigated in these very samples using identical methods, Filifactor is the third most prevalent for GAP and second most prevalent for CP patients and is thus at eye level with organisms that are considered key players in periodontal disease. At the same time, F. alocis shows the lowest prevalence in the PR group of all analysed organisms. Together with F. nucleatum, F. alocis is the only organism to show a significantly higher detection frequency in both GAP and CP patients compared to the PR group.
Using PCR-based identification methods may introduce bias, since structurally different organisms could exhibit different copy numbers of ribosomal genes and will generally respond differently to DNA isolation and the chosen set of broad range bacterial primers . However, the relevance of F. alocis is supported by several other epidemiological studies conducted in the past years using DNA-based techniques. F. alocis was detected in GAP patients as well as in CP patients with prevalence rates varying between 45%  and 90% , depending on the methods employed. Some authors propose F. alocis as a marker organism for periodontal disease  and even for the shift from periodontal health to disease .
Our data strongly support the findings of these studies and motivated the attempt to visualize F. alocis within the periodontal biofilm of GAP patients using FISH. The organism could be detected in high numbers in the majority of the examined carriers. The percentage of positive patients approximately matches the dot blot results. Strikingly, several areas of the biofilm show F. alocis in densely packed groups (Figure 4c) or as a part of concentric bacterial agglomerations (Figure 5d) - formations that suggest a certain degree of organisation to the observer. Moreover, the organism could be visualized in structures that are considered characteristic architectural features of periodontal biofilms. F. alocis is among the bacteria in mushroom-like protuberances on the surface of the biofilm (Figure 5b) and it contributes, grouped around what might be diffusion or convection channels, to the formation of structures reminding of test-tube brushes (Figure 5c). The close colocalization of F. alocis with other periodontal pathogens suggests that Filifactor might be involved in coaggregation events that take place during the establishment and maturation of the biofilms and that are thought to play a crucial role in biofilm formation . Moreover, the tight colocalization might indicate necessary symbiotic relationships that could help to explain the fastidiousness of Filifactor.
Just like group I treponemes , F. alocis predominantly colonizes the apical and middle third of the carriers and could only casually be detected in the cervical third. Most interestingly, the organism preferably settles on the side of the carrier facing the soft tissues and is thus in immediate contact to the host's immune defence. All these observations point to a causal involvement of F. alocis in the formation and maintenance of the analysed biofilms.
However, one might question whether these carrier-borne biofilms accurately model the unperturbed biofilms in periodontitis patients. Wecke et al.  compared the bacterial load after 3 and 6 days and showed that the biofilm mass covering the carriers increases with time. The presence of F. alocis on only one side of the membranes is further evidence that these samples are not simply fragments of biofilm torn out of the pocket during the removal of the carriers, but in fact newly grown biofilms that form while the carriers are in situ. Although FISH reveals structural elements specific to periodontal biofilms, one cannot deny that the introduction of the carrier into the periodontal pocket creates an artificial environment. The barrier between root surface and pocket epithelium might hamper access of the immune system to the bacteria on the tooth side, while only the biofilm growing on the soft tissue side actually faces the host. Moreover, these biofilms do not form on natural substrate but instead on ePTFE membranes. However, it seems likely that the substrate is of minor importance to the biofilm development. Wecke et al.  did not observe differences between biofilms grown on different carrier materials, and it is likely that the acquired pellicle, which covers both the root and the membrane, renders colonization conditions on a broad range of materials alike. This claim is supported by microscopic examination of the biopsy submitted to FISH. F. alocis could be visualized in high numbers and detected in arrangements similar to those seen in carrier-borne biofilms. Thus, a contribution of Filifactor to the structural organisation of 'naturally' grown biofilms seems highly probable.
The applied carrier system proves to be a valuable tool for the exploration of periodontal biofilms as it allows to investigate topographic relations within the pocket without invasive treatment. Subsequent FISH permits to analyse the distribution and colocalization of potential pathogens within the biofilm and can thus contribute to a better understanding of the complex host-microbe interactions that lead to periodontal destruction.
The prevalence of Filifactor alocis in both GAP and CP patients was found to be elevated as compared to PR control. F. alocis thus seems to be a powerful diagnostic marker organism for periodontal disease. FISH revealed the involvement of F. alocis in numerous structural arrangements that point to its potential role as one of the architects of structural organisation within periodontal biofilms. Filifactor alocis should be considered an important periodontal pathogen and warrants further research.
We thank Eva Kulik, University of Basel, and Eivind Strøm, University of Oslo, for providing clinical samples, Cindy Hefenbrock and Marie Knüver for excellent technical assistance, Derek Ramsey for proof reading, and Dr. Wolf-Ulrich Klotz for his support.
This work was supported by the Sonnenfeld-Stiftung, Berlin, Germany, and by a Rahel-Hirsch grant from Charité - Universitätsmedizin to AM.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.