Oral biofilms form part of the healthy oral environment and there is a well-balanced equilibrium between bacteria and the host. Two most common infection-driven diseases of the oral cavity, periodontitis, and caries, are the outcomes of the disruption in this equilibrium. Periodontitis is a periodontal pathogen-induced disease, where chronic inflammation initiates degradation of tooth-supporting tissues, including the gingival epithelium, periodontal ligament, and alveolar bone (Darveau, 2010). Caries occurs when the acid metabolites of cariogenic bacteria dissolve dental structures: enamel and dentin. As a matter of fact, the oral cavity acts as a reservoir of bacterial pathogens for systemic infections and several oral microorganisms have been linked to systemic diseases, i.e., Actinomyces species to brain abscesses (Könönen and Wade, 2015), Parvimonas micra to septic arthritis (Baghban and Gupta, 2016), and Streptococcus mutans and Porphyromonas gingivalis to atherosclerotic plaques (Fernandes et al., 2014).

The oral cavity is a dynamic environment; eating, drinking, and even sleeping affect the pH, temperature, salivary flow, and nutrition. Depending on population dynamics and environmental factors, the relationship amongst bacteria in the biofilm could be symbiotic or competitive. Several studies have already indicated that oral bacteria can communicate with each other via quorum sensing molecules (Shao and Demuth, 2010; Sintim and Gürsoy, 2016). One important signaling mechanism for bacteria to sense environmental stress and respond adequately is secondary messenger molecules. In this context, cyclic dimeric guanosine 3′,5′-monophosphate (c-di-GMP), a bacterial secondary messenger molecule has become a molecule of high interest, since elevated concentrations of c-di-GMP regulate many processes affecting the initiation and maturation of bacterial biofilms (Valle et al., 2013). This occurs through the c-di-GMP-regulated production of exopolysaccharides, synthesis of extracellular adhesins, and fimbriae (Sondermann et al., 2012). C-di-GMP levels in bacteria are regulated through synthesis by diguanylate cyclases and degradation by phosphodiesterases (PDE). Diguanylate cyclases, which have a GGD/EEF domain (diguanylate cyclases can have GGDEF or GGEEF domain and GGD/EEF is a short hand for these two domains), make c-di-GMP via the condensation of two molecules of GTP. Degradation of c-di-GMP is regulated by PDEs, which contain EAL or HD-GYP domains. These metabolism proteins, together with c-di-GMP binding proteins, such as PilZ, modulate the intracellular perception of c-di-GMP (Ryjenkov et al., 2006). A recent complete genome analysis of oral biofilm-associated bacteria revealed that the genomes of some oral bacteria, including Capnocytophaga ochracea, Prevotella melaninogenica, S. mutans, Streptococcus gordonii, Fusobacterium nucleatum, Aggregatibacter actinomycetemcomitans, and Filifactor alocis, do not carry GGDEF or GGEEF, EAL, HDGYP, or PilZ domains, while some other oral bacteria, such as Treponema denticola and Selenomonas sputigena, do (Römling et al., 2013).

Recently, another bacterial second messenger, c-di-AMP has emerged as an important regulator of bacterial biofilm formation. This messenger molecule has a regulatory role in cell wall homeostasis, fatty acid synthesis, and biofilm formation (Opoku-Temeng et al., 2016). C-di-AMP is synthesized by diadenylate cyclases and degraded by PDEs. Several c-di-AMP receptors in various bacteria have been detected lately (Opoku-Temeng et al., 2016). However, there is a paucity of literature reports that have documented the presence of proteins containing diadenylate cyclase domains in oral bacteria.

Cyclic dinucleotide signaling in oral bacteria has been under-studied and, hence, there is almost no information about how cyclic dinucleotides affect oral bacterial physiology and virulence. This perspective highlights the dearth of cyclic dinucleotide research in this important class of bacteria.

Many methods have been developed to detect c-di-GMP in both cell lysates and in live bacteria. A high performance liquid chromatography and an ultra-performance liquid chromatography-tandem mass spectrometer are the methods of choice for a sensitive detection of c-di-AMP and c-di-GMP in bacterial cell lysates, but they both are inappropriate for live imaging of cyclic dinucleotides. A competitive enzyme-linked immunosorbent assay was developed as a simple method for the quantification of c-di-AMP levels in bacteria, however, the minimum detection limit (10 nm) of this method does not allow to measure c-di-AMP in culture media or in host cells (Underwood et al., 2014). Sintim and his co-workers demonstrated that low levels of c-di-GMP can be detected using a c-di-GMP riboswitch that was fused to a spinach aptamer (Nakayama et al., 2012). Later, the Hammond group adopted this design and modified a version of the riboswitch-spinach aptamer to detect c-di-GMP in live bacteria (Kellenberger et al., 2013). A limitation of this approach is the need to transform bacteria before c-di-GMP can be imaged.

Detection of intracellular cyclic dinucleotides in oral bacteria would allow us to understand their adaptive properties against a changing environment. However, oral bacteria live in multispecies biofilms where they have a continuous interaction, which potentially regulates their secondary messenger molecule formation and degradation. We tested this hypothesis in a dual bacterial co-culture model using F. nucleatum and four different selenomonads (Selenomonas noxia, S. flueggei, S. artemidis, and S. sputigena). F. nucleatum is a key species in dental biofilms by facilitating the colonization of periodontitis-associated pathogens by coaggregation-mediated mechanisms, and many Selenomonas spp. are known to coaggregate well with F. nucleatum (Kolenbrander et al., 1989). In our experiments, the four Selenomonas spp., i.e., S. flueggei (64 × 10⁷ CFU), S. sputigena (59 × 10⁷ CFU), S. noxia (64 × 10⁷ CFU), and S. artemidis (65 × 10⁷ CFU), were cultured, either alone or together with F. nucleatum (64 × 10⁷ CFU), in a reinforced clostridial medium at anaerobic conditions for 24 h. The bacterial cells were then centrifuged and the pellet was resuspended in 10 nm Tris-HCl containing 100 mM NaCl. Lysozyme (L6876, Sigma-Aldrich, St. Louis, USA) was added with a final concentration of 62.5 μg/ml. The samples were resuspended, sonicated, and heated to 90°C for 30 min, centrifuged once again, and extracted with CH₃OH:CH₃CN:H₂O. The supernatants were centrifuged with Ultracel 3K centrifuge tubes (Amicon® Ultra-4 Centrifugal Filter Units, Merck Millipore, Germany) for 40 min at 4,000 g at a room temperature, and evaporated to dryness. Nucleotides were extracted and c-di-GMP levels were measured with ultra-performance liquid chromatography—tandem mass spectrometer.

The numbers of bacterial cells in each group at the end of the experiment are given in Figure 1. According to our results, S. sputigena produced highest levels of c-di-GMP. Surprisingly, when co-cultured with F. nucleatum, c-di-GMP levels decreased nearly with 50%. This was an interesting finding, since it was expected that the co-culture of S. sputigena with F. nucleatum will increase its coaggregation and, hence, biofilm formation. However, reduced c-di-GMP levels indicate an increased motility and decreased biofilm formation (Figure 1).

To date, the reported technologies for detecting of c-di-GMP in live bacteria have not been powerful or simple enough to use for the simultaneous detection of c-di-GMP in polymicrobial biofilms. The ultra-performance liquid chromatography-tandem mass spectrometer is a post sampling method and the spinach-riboswitch method requires transfection, which may not work with all bacteria. Secondly, it is unclear whether the biology that emerges from a transfected bacterium represents what happens in a native scenario. The development of easy-to-use and broadly applicable c-di-GMP and c-di-AMP detection methods for various oral bacteria enables the delineation of how cyclic dinucleotides affect signaling in oral biofilms.

Disruption of biofilms by using molecules that interfere with cyclic dinucleotide regulation is a promising approach to fight against infectious diseases. In general, enhanced c-di-GMP levels intracellularly lead to biofilm formation, so it is expected that activators of PDE, inhibitors of diguanyl cyclases, or the supramolecular aggregation of the signaling molecules would inhibit formation (Kelsey et al., 2012; Nakayama et al., 2016). In this context, exogenous PDE was demonstrated to disperse Pseudomonas aeruginosa biofilms and reduce its biofilm formation (Hickman et al., 2005). Interestingly, the inhibition of RocR PDE in P. aeruginosa did not lead to enhanced biofilm formation but rather swarming inhibition, indicating that there are nuances to cyclic dinucleotide signaling and the interception of signaling with small molecules (Zheng et al., 2016). A few inhibitors of diguanylate cyclases have been identified or developed, and these were found to reduce bacterial biofilms (Opoku-Temeng et al., 2016). For example, diguanyl cyclase inhibitors can reduce Vibrio cholerae or P. aeruginosa biofilm formation (Sambanthamoorthy et al., 2012, 2014).

C-di-AMP is involved in biofilm formation as well as cell wall synthesis in gram-positive bacteria (Opoku-Temeng et al., 2016). The Sintim group has identified several small molecules that inhibit c-di-AMP synthases (Zheng et al., 2014; Opoku-Temeng and Sintim, 2016a,b) and some of these compounds inhibit the growth or biofilm formation of S. aureus (Opoku-Temeng et al., 2017).

Cyclic dinucleotides were discovered already 30 years ago. During the past decade, a renewed interest in these fascinating second messengers led to a tremendous progress in our understanding of how cyclic dinucleotides regulate bacterial lifestyle but also bacteria-host interactions. Oral microorganisms form some of the most intricate biofilms, yet c-di-GMP and c-di-AMP signaling, which regulate biofilm formation, have been rarely studied in oral bacteria. Perhaps this dearth of information on dinucleotide signaling in oral bacteria is due to initial bioinformatics predictions, stating that many oral bacteria do not harbor proteins that are capable of making or degrading cyclic dinucleotides. Recent, but still few reports, however, challenge this perception, and c-di-GMP or c-di-AMP signaling has now been recognized from some oral bacterial species. We encourage the scientific community to take a second look at cyclic dinucleotide signaling in putative oral pathogens, as there might be potential drug targets in dinucleotides that could be usable in preventing oral biofilm-related diseases like periodontitis.

UG, MG, EK, and HS: Substantial contributions to the conception or design of the work, drafting and critical revision of the article, and final approval of the version to be published.