Periodontal infections, or periodontal diseases, are a set of chronic inflammatory diseases that destroy the tooth-supporting (periodontal) tissues. They are caused by oral bacterial biofilms attaching on the tooth surface. They have the capacity to trigger a series of inflammatory responses, which may destroy the gingival tissue and the alveolar bone supporting the tooth, if they become exacerbated [1],[2]. With regards to its capacity as an ecological niche, the oral cavity can be colonized by more than 700 species [3] and approximately 500 of those can be present within the forming biofilms [4],[5]. Among the biofilm-associated microbiota, earlier clinical epidemiological studies have demonstrated that three species in particular, also designated as the red complex, are more associated with periodontal disease than others. These are namely Porphyromonas gingivalis, Tannerella forsythia, and Treponema denticola. They are all Gram-negative anaerobes, with a high proteolytic activity [6]. Among these three, P. gingivalis holds a prominent role in orchestrating the virulence of the biofilm and the consequent tissue inflammatory response, earning itself the characteristics of a keystone periodontal pathogen [7],[8]. P. gingivalis expresses several virulence factors, including, fimbriae, LPS, and its cysteine proteases, namely gingipains [9]. These include the arginine-specific proteinases RgpA and RgpB, and the lysine-specific proteinase Kgp, which represent the majority of the cell-surface proteinases of P. gingivalis[10]. Clinical studies have demonstrated that periodontal infection associated with P. gingivalis can result in significantly elevated systemic antibody response to the gingipains [11],[12].

When growing in a subgingival (below the gingival margin) biofilm under strict anaerobic conditions, P. gingivalis is highly dependent on its gingipains for utilizing free amino acids as a source of carbon and nitrogen [13]. Moreover, unlike other gram-negative bacteria, P. gingivalis does not produce siderophores to sequester and transport iron but its gingipains mediate the uptake of iron from hemoglobin, heme proteins, and ferritin [14],[15]. Gingipains are also considered important in the capacity of P. gingivalis to evade host defences, by degrading antibacterial peptides, such as neutrophil-derived α-defensins, complement factor, such as C3 and C4, T cell receptors, such as CD4 and CD8 [16]. Alternatively, P. gingivalis and its gingipains can subvert the host immune response by proactively manipulating host molecules, particularly of the complement [17],[18]. For instance, P. gingivalis may perturb the cross-talk between C5a receptor and toll-like receptor signalling in order to prevent bacterial clearance and cause dysbiosis [19], eventually resulting in periodontal bone loss [20],[21]. The construction and phenotypic analysis of isogenic protease mutants of P. gingivalis have confirmed putative functions for these proteolytic enzymes [22]. In vivo studies using the P. gingivalis mutant strains in animal models have reinforced the view that the gingipains can modulate the infection process [23]-[26]. In vitro studies have demonstrated an involvement of the gingipains in the regulation of inflammatory mediators from various host cells, including IL-1 α, IL-1β, IL-18 [27], receptor activator of NF-κB ligand (RANKL) [28]-[31], tumor necrosis factor-α converting enzyme (TACE) [32], protease-activated receptor (PAR)-2 [33], or soluble triggering receptor expressed on myeloid cells (sTREM)-1 [34].

Understanding how different organisms act within a given polymicrobial biofilm brings us closer to understanding the etiological mechanisms of periodontal disease [1]. That is because interactions among different bacterial cells can determine the structural characteristics, maturation and virulence of the biofilms [35]-[37]. These interactions can occur at several levels, including physical contact, metabolic exchange, and signal-mediated communications [38]. Additionally, species-specific virulence factors may regulate bacterial growth, hence altering the conditions of the ecological niche for biofilm formation. In this respect, most studies involving gingpains have focused on P. gingivalis as a single species, which might overlook the bacterial interactions within a complex biofilm community. Therefore, the present study used a 10-species subgingival biofilm, aiming to investigate the role of gingipains on the growth and structure of the biofilm, by incorporating P. gingivalis gingipain-deficient strains.

As it is well established that periodontal diseases are initiated by a mixed-species biofilm [39],[40], in vitro biofilm models, may be more accurate in studying the causative factor of the disease, than single species in planktonic form [37],[41],[42]. The present study investigated the involvement of P. gingivalis gingipains in the quantitative and qualitative composition of a polymicrobial biofilm consisting of 10 species that are frequently comprising part of the subginvival microbial flora. Among their many properties, gingipains are important for the growth of P. gingivalis and as transporters for iron [14]. While in planktonic culture P. gingivalis gingipain deficient strains require longer doubling times [43], their incorporation into a polymicrobial biofilm did not yield differences in numbers, compared to the wild-type strain. Hence, the growth characteristics of P. gingivalis may differ depending on whether it grows in planktonic or biofilm state. When present in a biofilm, gingipains do not appear to be crucial for the growth of P. gingivalis, as shown here. Interestingly, among the remaining nine species in the biofilm, the only one whose growth was affected by the presence of gingipains was T. forsythia. In particular, the P. gingivalis Lys-gingipain deficient strain resulted in a strong reduction in T. forsythia numbers after 64h of biofilm growth. Reversely, this indicates that the Lys-gingipain produced by P. gingivalis has an additive effect on the growth of T. forsythia in the biofilm. This denotes a synergistic association between T. forsythia and P. gingivalis as mutual components of a polymicrobial community, which is mediated by the Lys-gingipain of the latter.

Previous studies have shown that gingipains are crucial for the co-aggregation of P. gingivalis or its co-adhesion with other species, such as T. denticola[44]-[46], or for the invasion of host cells [47]. Hence, the gingipains may not only affect the quantitative composition but also the structural conformation of the biofilm. For this reason, the biofilm architecture was also investigated by CLSM. P. gingivalis occurred in distinguishable and evenly distributed clusters within the biofilm regardless of whether it expressed a gingipain or not. The communities of T. forsythia within the biofilm exhibited similar patterns to those of P. gingivalis, and were frequently co-localized, yet without impinging onto each other. The proximal association of these two species communities in biofilm may hint for an ecological relationship. This is also substantiated by the notable absence of T. forsythia clusters from the vicinity of the Lys-gingipain deficient P. gingivalis. Hence, this gingipain may be important for the growth of T. forsythia and its spatial interdependency to P. gingivalis within the biofilm. This observation could represent an example of the metabolic responses and bacterial quorum-sensing within the biofilm [48].

Another interesting observation of the present study is that of the structural re-arrangement of T. denticola in the biofilm, depending on the presence or absence of the Arg-gingipain. Earlier studies have shown that other species can interact with P. gingivalis in both planktonic suspensions and biofilms [46],[49],[50]. A recent study using the similar multi-species biofilm model as here demonstrated that P. gingivalis and T. denticola have the tendency to co-colonize gingival epithelial tissue [51]. In a dual P. gingivalis - T. denticola biofilm, it was also demonstrated that gingipains do contribute to their interaction [50]. In the present experimental model, T. denticola cells formed dense circular clumps with the wild-type P. gingivalis strain. However, in the presence of the P. gingivalis Arg-gingipain deficient strain, this conformation was lost and T. denticola cells were instead arranged in looser threaded structures, even though their numbers in the biofilm were not changed. This finding provides further evidence of the ecological association between P. gingivalis gingipains and the structural arrangement of T. denticola in the biofilm. It is difficult at this stage to interpret the biological meaning of this change in T. denticola structure. Of note, in a recent study using the similar biofilm model it was demonstrated that omission of streptococci from the biofilm resulted in numeric changes of P. gingivalis and P. intermedia. The latter also lost its aggregated form and was arranged in filamentous long chains, resembling those of the missing streptococci [35].

This study showed that the gingipains of P. gingivalis promote quantitative and qualitative shifts in the composition and structure of a multi-species biofilm. More specifically, the Lys-gingipain enhances the growth of T. forsythia, whereas the Arg-gingipain promotes the aggregation of T. denticola in the biofilm. These ecological interactions are interpreted as synergistic ones, and may support the survival and the virulence of the biofilm community as a whole.

NB and GNB conceived the study. BK, GNB, TT, JAO, MAC and NB designed the study. JAO and MAC generated and provided the Porphyromonas gingivalis gingipain mutants. BK performed the experiments. BK and NB performed the data analysis. BK, GNB and NB wrote the paper. TT, JAO and MAC reviewed and approved the final version of the paper. All authors read and approved the final manuscript.