Towards microbiome transplant as a therapy for periodontitis: an exploratory study of periodontitis microbial signature contrasted by oral health, caries and edentulism

Potential for the periodontitis bacteriotherapy

The first step towards developing a bacteriotherapy for periodontitis is to establish
the existence of a microbial community characteristic of periodontitis as well as
the community characteristic of oral health. Two additional communities from edentulous
and caries patients, whose microbial communities are expected to be distinct, were
added for contrasting purposes.

Several studies have compared microbial communities isolated from different sites
in the oral cavity and from patients with different clinical conditions (e.g., 14], 15], 17], 19], 44]–47]). One of these studies revealed distinct partitioning of bacterial communities in
subgingival biofilms from healthy and diseased periodontal sites 45]. Another study indicated that not only the oral microbial species were drastically
different between periodontal health and disease but also their gene expression 48]. Moreover, in subjects with periodontitis, disease-associated bacteria were found
on the buccal mucosae, tongue, and saliva indicating that the entire oral microbiome
might be altered 49]–53].

Since the bacteriotherapy would likely be performed by attempting to transplant both
sub- and supra-gingival microbes, the present study was set to test the hypothesis
that overall compositions of the oral microbial communities were distinct between
the conditions of periodontitis, edentulism, caries, and oral health. In fact, the
entire oral microbiome of each subject was obtained by pooling together several oral
cavity sub-microbiomes such as subgingival, supragingival, and mucosal membranes.

The key finding of this study, critically important for the potential bacteriotherapy,
is that microbial compositions (i.e., signatures) of the entire (or substantial subset
of) oral microbiomes of subjects with periodontitis contrasted by edentulism and oral
health are distinct. In addition, the microorganisms which are usually considered
as pathogens of periodontitis, P. gingivalis, T. denticola and T. forsythia, are not the sole representatives of the periodontitis microbial signature. In particular,
this study provides evidence that a broad microbial-community model of oral health
may offer a valuable perspective when considered in conjunction with (or in contrast
to) more common single-pathogen models (7] and refs. therein). With this perspective, it is reasonable to expect that microbial
community composition is a function of interspecies competition and cooperation, which
provides optimism for further developing the bacteriotherapy. Below, we expand on
the motivation for choosing molecular methods and sampling procedures.

Methodological validity

Historically, subgingival oral microbiome was predominantly assessed by a checkerboard
hybridization technique revealing signals from approx. 40 species (for review, see
54]). Pathogenicity of a few species, i.e., “colored complexes”, was presumed to account
for oral conditions, such as periodontitis. With the advent of next generation sequencing,
extensive cataloging of the subgingival microbiome has been conducted (e.g., 17], 45], 55]). In our study we also utilized high-throughput sequencing technology, because it
provides an in-depth qualitative and quantitative characterization of the microbial
communities, which is essential for discovery of microbial signatures.

The data for the analysis of microbial signatures were the relative abundances of
microorganisms measured by the number of sequencing reads obtained from high-throughput
sequencing of each subject’s oral microbiome. Contrary to what one might expect, it
is important to note, that absolute abundances are not attainable by the high throughput
sequencing due to the biases in amplification and other physicochemical reasons 56], 57]. After sequencing reads are processed, each subject is represented by a p-dimensional
vector (p?=?597) in the multidimensional space, with each coordinate corresponding
to the relative abundance of each microbial species. The ordination analysis 58] presented in the Results section is a standard dimension-reducing method, which reveals
the structure of the 3 groups (Fig. 4). Interpretation of these groups is based on the subjects’ microbial abundance profiles
that are most similar and hence grouped closer to each other. Although the sample
size is rather small (n?=?16), the quality of the clusters and their statistical significance
(based on permutation) suggests that the observed grouping may be reproducible in
a larger, independent cohort. Other researchers, e.g., Kumar et al. 47], showed with a larger group that there is a clear separation between microbial communities
for healthy and periodontitis and peri-implantitis patients. Another potential issue
with generalization of our findings could be due to the fact that all health subjects
were from Europe. Nevertheless, as was discovered by Nasidze et al. 59], the compositions of the oral microbial communities obtained from 12 worldwide locations
of 10 individuals each were “larger among individuals from the same location than
among individuals from different locations”. Hence the differences in the composition
among individuals were not determined by geography.

Periodontitis pathogenesis: “red complex” or entire community shift?

This study revealed grouping of subjects based on the composition of their entire
oral microbiomes, which is consistent with previously reported results based on the
microbial composition of their subgingival plaque samples 47]. Specifically, it has been suggested that there exist distinct microbial signatures
characteristic of periodontitis, peri-implantitis, and oral health. Our research expands
on these findings by investigating the entire oral microbiome including combined together
subgingival, supragingival, and mucosal sub-microbiomes. Building upon a recent report
by Griffen et al. 45], which showed a high microbial diversity of sub-gingival biofilms in subjects with
periodontitis compared to oral health, our study also revealed a similarly increased
diversity in the entire oral microbiome of subjects with periodontitis compared to
edentulism, oral health and caries. Among the four conditions, our study showed the
lowest microbial diversity in oral microbiomes of subjects with edentulism. This is
presumably due to the absence of teeth and subgingival pockets in which otherwise
mature biofilms develop 45]. It is important to note that there are other studies focused on microbial signatures
of periodontitis (e.g., 60]), however these studies did not attempt to contrast several conditions.

Our observations support the polymicrobial nature of periodontitis contrary to the
“old-school” long-standing notion of specific pathogens believed to cause periodontitis.
Indeed, the oral microbial signature in periodontitis subjects included P. gingivalis, T. denticola and T. forsythia, which are collectively referred to as the “red complex”. As our results showed,
red complex bacteria were not the absolute determinants of the periodontitis microbial
signature. The presence of the red complex bacteria in oral health and other conditions
could be attributed to differences in: (i) the host immune response, (ii) genes encoding
virulence factors among different strains, and (iii) gene regulation 51]. The findings of distinct oral microbial signatures between subjects with periodontitis
and oral health supported the notion that periodontitis is associated with a dysbiosis
(i.e., microbial imbalance) of the entire oral microbiome. The oral dysbiosis may
be due to the assembly of a synergistic microbial community that tilts the balance
away from a health-associated microbial homeostasis 51]. It could be initiated by the acquisition of virulence factors by microbes and/or
changes in the regulation/expression of genes within the community 17], 61]. Perhaps one or several microorganisms, e.g., P. gingivalis, are capable of initiating synergetic interactions that ultimately result in the
emergence of disease-provoking microbial community 7], however a mechanistic proof is lacking.

The idea for the potential bacteriotherapy of periodontitis is inspired by the success
of the fecal microbial transplant treatment of Clostridium difficile (CD) infection. It should be noted that while the diversity of bacteria associated
with periodontal disease is very high 55], 62], the diversity of bacteria associated with CD infections is very low 63]. The reason to emphasize these differences is that the goal of bacteriotherapy for
CD is to increase bacterial diversity whereas the goal of bacteriotherapy for periodontitis
is not the same since the bacterial diversity is already high. Presumably in both
diseases, replacement of healthy stable microbial communities is the answer to prevent
future dysbiosis. However, without performing actual experiments, one cannot speculate
on the success or failure of bacteriotherapy for periodontal disease.

Antimicrobial treatment before the transplant

It is reasonable to assume that for the bacteriotherapy to be successful, the recipient’s
oral cavity has to have lowest possible load of microorganisms to allow the new microbiome
to flourish without interference from the previous disease-associated microbiome.
Administration of adjunctive systemic antibiotics 64], 65] or a full-mouth disinfection approach that reduces the bacterial reservoir in all
habitats of the oral cavity will result in more pronounced shifts in the microbial
composition and significantly reduces the incidence of clinical attachment loss compared
to supra- and subgingival debridement alone 37], 66], 67]. Still, it is desirable to administer such an antimicrobial agent that is effective
against a broad spectrum of microbes and is easy to neutralize with another nontoxic
agent, because as most antiseptics remain active in the oral cavity for quite some
time 68], 69] and therefore, may affect the viability of an oral microbial transplant. We have
been able to demonstrate that sodium hypochlorite (NaOCl), which is bactericidal against
a panel of oral biofilm microorganisms 68], 70], 71], can be inactivated by a nontoxic sodium ascorbate – ascorbic acid buffer. Although
some critics have said that NaOCl treatment is antiquated and dangerous, a recent
study proved the opposite 72]. Moreover, the Galvan et al. study 72] showed safety and a significant reduction of plaque and bleeding on probing in the
group that used NaOCl rinses compared to the control group.