Dental implants have become the primary protocol for the treatment of dentition defects and edentulous patients nowadays. Despite its high success rates, complications could reduce the long-term cumulative implant survival (Fu and Wang, 2020). Peri-implantitis (PiM), as the primary cause leading to implant failure, could arise about 22% in 10 years after implantation (Derks and Tomasi, 2015). During PiM pathogenesis, the inflammation at the implant-bone interface is evoked to successively cause bone resorption, gingival tissue retraction, and implant surface exposure. To date, plaque biofilm is a putative and pivotal factor that induces the initial inflammation of PiM (Berglundh et al., 2018). Hence, eliminating biofilm is essential for the prevention and treatment of PiM.

In clinical practice, methods for decontaminating the implant surface could be divided into physical and chemical approaches. Physical approaches, such as mechanical scaling, sandblasting, magnetic nanoparticles, and laser therapy, could damage the biological surface, cause material residue, or be difficult to eliminate the massive biofilm matrix enmeshed in the micropores due to the rough character of the implant surface (Elbourne et al., 2020; Lasserre et al., 2020; Vyas et al., 2020). Correspondingly, chemical approaches, including antimicrobial agents, antimicrobial peptides, and photodynamic therapy (PDT), et al., mainly focus on eradicating bacteria or disturbing biofilm’s integrity (Chen and Lee, 2018; Liu et al., 2019; Pinto et al., 2020). Whereas the biofilm extracellular matrix, mainly composed of exopolysaccharides (EPS), lipids, proteins, and extracellular DNA (eDNA), represents around 90% of the total biofilm biomass (Fulaz et al., 2019). Consequently, even if the bacteria were thoroughly damaged, the massive biofilm matrix remains attached to the implant surface. Nevertheless, the extracellular matrix protects bacteria from agents penetration and shear stress, making a low bioavailability of agents’ germicidal effects (Flemming and Wingender, 2010). Previous in vivo studies have demonstrated that there was fibrous connective tissue between regenerated bone and exposed implant surface with PiM after grafting autogenous bone grafts (You et al., 2007), indicating that biofilm-associated contaminants obstructed the re-osseointegration of the implant surface. Till now, the methods for treating PiM are unpredictable yet. Thus, strategies for eradicating the whole biofilm should be prospective.

The biofilm extracellular matrix constitutes the main component of biofilm, while EPS are one of the most constituents of the extracellular matrix (Pinto et al., 2020). Hence, EPS is indispensable to biofilm formation and constitutes the protective barrier of encapsulated bacteria (Flemming and Wingender, 2010). Besides, EPS occupies the main adhesion force binding to the biological surface as well as bacteria and proteins through Van der Waals force, electrostatic attraction, hydrogen bonds (Arciola et al., 2018), thus facilitating the cohesion of the biofilm structure. In addition, EPS is responsible for water retention within the biofilm, and provides nutrient sources and ions for the inner bacteria (Pinto et al., 2020). Therefore, EPS could be a potential target for eradicating the biofilm on implant surface.

Similar to plaque biofilm on teeth, the biofilm formation on implant surface possesses typical stages, namely initial, early, secondary, and late colonizations (Vilarrasa et al., 2018; Bermejo et al., 2019). Streptococcus oralis and Actinomyces viscosus (A. viscosus) are the representative strains in the initial stage, then Veillonella parvula in the early stage, Fusobacterium nucleatum in the secondary stage, and Porphyromonas gingivalis in the late stage. In fact, Fusobacterium nucleatum and Porphyromonas gingivalis could only produce capsular polysaccharides but not EPS (Davey and Duncan, 2006), while Veillonella species utilize EPS secreted by Streptococci to adhere to biofilm (Liu et al., 2020). In addition, in the study of multispecies biofilms growing on the implant surface, Actinomyces showed about two-fold amount of biomass than Streptococcus (Bermejo et al., 2019). Moreover, Actinomyces plays an essential role as physical bridges to mediate coaggregation and coadhesion between coaggregating partners, which occurs 5- to 10-fold more often than coadhesion between noncoaggregating cells (Kolenbrander, 2000). Thus, disintegrating the Actinomyces EPS and the Actinomyces-associated biofilm is critical for the decontamination of the implant surface.

EPS could be composed of glucose, mannose, galactose, N-acetyl-glucosamine, and other monosaccharides (Rabin et al., 2015), which could link as α-1,4 bond, β-1,4 bond, or β-1,3 bond. Glycoside hydrolases are enzymes that hydrolyze the glycosidic linkages between carbohydrates within polysaccharides or oligosaccharides (Naumoff, 2011). α-amylase (Amy) and cellulase effectively disrupted Staphylococcus aureus and Pseudomonas aeruginosa coculture biofilms by breaking down complex polysaccharides. However, whether glycoside hydrolases could disrupt A. viscosus biofilm remains unknown.

Our previous studies demonstrated that D-arginine (R) could disperse Porphyromonas gingivalis mature biofilm in 72 h (Li et al., 2020; Zhang et al., 2021). Whereas the treatment time takes too long for feasible clinical practice. The capability of R to disrupt A. viscosus biofilm and higher treatment efficiency should be further tested.

Therefore, the aim of this study was to investigate the effects of Amy on disassembling A. viscosus biofilm and the enhancing effects of R on Amy. Molecular dynamics (MD) simulations and molecular docking were furtherly exploited to explore the intrinsic mechanisms.

This study investigated the effects of R on enhancing Amy disassembling A. viscosus biofilm, and elucidated the theoretical molecular mechanisms. 8 mM R could prominently enable 0.01% Amy to disassemble A. viscosus biofilm in 30 min without cytotoxicity, and 8 mM R performed better effects on enhancing Amy hydrolyzing starch than 60 ppm Ca²⁺. Furtherly, the mechanisms for R enhancing Amy include R increased the overall stability of Amy and the activity of the catalytic triad, and R also increased the stability of calcium-binding sites.

Implants have been widely used in oral rehabilitation, artificial joints, bone fixators, and other bone-related fields (Hanawa, 2019). An untainted titanium surface is the guarantee of osseointegration (Buser et al., 2017). Nevertheless, biofilm-associated infections are the main cause of the early failure after implantation or for the PiM during the long-term period. Unlike other application fields, dental implants are threatened by bacterial infection for their whole period due to their contact with the multimicrobial oral environment. Plaque biofilm will accumulate around implant attributed to bad oral hygiene habits, food impaction, smoking, et al. Subsequently, the immune response is evoked, leading to soft tissue inflammation, bone resorption, and implant surface exposure (Fu and Wang, 2020). Hence, it is essential to eliminate biofilm to terminate the process of PiM and promote the regeneration of bone around the implants.

Eradicating bacterial biofilms has been a huge concern nowadays. However, the administration of conventional antimicrobial agents lacks efficiency since they fail to approach the massive bacteria encapsulated by the extracellular matrix (Koo et al., 2017). Thus, many studies reported effective antibiofilm agents capable of disrupting the extracellular matrix. Nanocarriers, designed positively charged, could be more likely to interact with the extracellular matrix (Fulaz et al., 2019). For instance, chitosan oligosaccharide-capped gold nanoparticles could eradicate mature Pseudomonas aeruginosa (P. aeruginosa) biofilm by electrostatic interactions (Khan et al., 2019). On the contrary, tobramycin, being positively charged, was blocked by the P. aeruginosa biofilm, while the neutrally charged ciprofloxacin could easily penetrate the biofilm (Tseng et al., 2013). These contradictory results indicate that nanocarriers could not penetrate biofilms with the pure electrostatic theory. Moreover, recent strategies of nanocarriers mainly focus on disrupting eDNA or proteins, not the more quantities of polysaccharides. PDT involves the use of photosensitizers, leading to the production of reactive oxygen species (ROS). ROS is considered to oxidize the cellular components, like lipids and DNA. In fact, ROS could also have the capability of attacking extracellular matrix molecules by disrupting the cell-to-cell and cell-to-matrix interactions, which causes the degradation of matrix structure (Li et al., 2013). Nonetheless, PDT also possesses the difficulty to penetrate the biofilm matrix easily and could not degrade the complex matrix rapidly (Pinto et al., 2020). Enzymes that possess the capability of degrading EPS, protein or eDNA could be employed as potential biofilm disruptive agents. Dispersin B is a well-known glycosyl hydrolase that could disrupt the major EPS of Staphylococcus epidermidis biofilms by specifically hydrolyzing poly-N-acetylglucosamine (Chen and Lee, 2018; Piarali et al., 2020). Deoxyribonuclease I (DNase I) is highlighted for antibiofilm purposes due to its effects on disrupting eDNA. For instance, DNase I efficiently increased vancomycin activity against the biofilm of Enterococcus faecalis, which could decrease the dosage of vancomycin by 8-fold (Torelli et al., 2017). In the present study, Amy showed significant effects on dispersing A. viscosus biofilm and hydrolyzing starch with the addition of R, implying that the Amy + R might be an optimal compound for decontaminating the biofilm-contaminated implant surface.

The PiM-associated biofilm is a cluster of definite bacterium complex. Several studies have assessed bacterial adhesion and colonization on implant surfaces through simple biofilm models or multispecies biofilm models (Schmidt et al., 2017; Vilarrasa et al., 2018; Bermejo et al., 2019). Despite the diversity of bacterial strains, Actinomyces is the widely used species for its initial adherence on titanium surfaces as early as 15–20 min after incubation (Guggenheim et al., 2001; Schmidlin et al., 2013). Similarly, Actinomyces was selected in the present study to represent the most critical species in forming the biofilm of PiM. After 48 h of incubation, the biofilm expressed compact and multi-layered (Figure 6), consistent with the study of Yamane et al. (2013). Especially, there were porous micro-pores among the biofilm, which might facilitate the exchange of nutrients and gas. A previous study demonstrated that the major components of the EPS of A. viscosus were: N-acetylglucosamine (62%), galactose (7%), glucose (4%), uronic acid (3%) (Rosan and Hammond, 1974). Whereas another study revealed that 39% galactose, 37% N-acetylglucosamine, 19% glucose, and 5% mannose were the main components of EPS produced by A. viscosus (Imai and Kuramitsu, 1983). Theoretically, N-acetylglucosamines compose poly-N-acetylglucosamine by forming β-1,6 glycosidic bond, galactoses compose lactose by forming α/β-D-galactose residues, and glucan, starch, as well as cellulose could be composed of glucose either by α-1,4 glycosidic bond, or α-1,6 glycosidic bond or β-1,4 glycosidic bond. Therefore, in order to find the feasible glycoside hydrolases for hydrolyzing the above-mentioned glucosidic linkages, Dispersin B, α-galactosidase, β-galactosidase, dextranase, cellulase, and Amy were employed. Unexpectedly, merely Amy at supersaturated concentrations showed assumed results. Amy specifically endohydrolyzes the α-1,4-D-glucosidic linkages in polysaccharides containing three or more α-1,4-linked d-glucose units. Hence, we presume that the EPS of A. viscosus biofilm are mainly composed of polysaccharides containing α-1,4 glycosidic bond, which is inconsistent with previous studies (Rosan and Hammond, 1974; Imai and Kuramitsu, 1983). Fleming et al. showed that 0.25% Amy reduced the biomass of Staphylococcus aureus and Pseudomonas aeruginosa polymicrobial biofilms (Fleming et al., 2017). However, the saturation concentration of Amy is merely 0.01%, higher concentrations could lead to apparent cytotoxicity (Figure 2). As a consequence, 0.01% Amy was furtherly investigated.

Most thermostable amylases require the additional Ca²⁺ for their thermostability. For instance, 5 mM Ca²⁺ enhanced the relative activity (%) of α-Amy from 100 to 115% (Lin et al., 1998). 1, 5 and 10 mM Ca²⁺ enhanced the relative activity of α-Amy from 100 to 105, 109, and 116%, respectively (El-Banna et al., 2007). The mechanism might be that the binding of Ca²⁺ ions to the α-helical structure of α-Amy increases the overall stability of α-Amy. Figure 4D showed that the mean relative activity of 0.01% Amy was raised from 100 to 113.0% with the addition of 60 ppm Ca²⁺, but the statistical difference was scant. Strikingly, 8 mM R dramatically increased the efficiency of Amy to 175%. Moreover, 60 ppm Ca²⁺ and 8 mM R showed a synergistic effect on enhancing the relative activity of Amy from 100 to 240%. These results indicate that R expresses a better effect than Ca²⁺ on enhancing Amy, and there is a synergistic mechanism between them. Figure 8F depicts the three-dimensional structure at the calcium-binding regions after MD simulations. There were massive hydrogen bonds between the residues of Amy and R. The formally defined calcium-binding sites also showed strong affinity with R. Additionally, the values of RMSF from PHE105 to ASN151 showed a sharp reduction with the addition of R (Figure 8C), which could be attributed to the firmly bonding interactions between R and Amy.

The results of CLSM are in line with SEM, both of them indicated that there remained few individual bacteria and sparse EPS after the treatment of 0.01% Amy + 8 mM R. Magdalena et al. reported that glycoside hydrolase (PelA_h) reduced the Pseudomonas. aeruginosa cells and polysaccharide elements, remained visible bacterial cells on the membrane (Szymanska et al., 2020). Similarly, minor A. viscosus could be observed on the plate in this study, whereas the remaining cells were bare without the enmeshing of extracellular matrix. Bacterium could adhere to biomaterials through capsular polysaccharides, fibronectin-binding proteins, collagen-binding adhesin, lipoteichoic acid, or other surface components (Arciola et al., 2018). Likewise, TEM showed the sparse extracellular matrix on the surface of the remaining cells (Figure 7). Hence, although the EPS of A. viscosus biofilm were entirely eliminated by Amy + R, the initial cells adhering to the surface could not be removed easily by Amy + R for their irreversible attachment via active adhesion (Carniello et al., 2018). It is worth noting that 0.01% Amy seemed to increase biofilm biomass (Figure 5C), which could be interpreted by the morphology of SEM (Figure 6). Briefly, after the treatment of 0.01% Amy, there was no notable reduction of biofilm, nevertheless, several cracks and huge cavities were created to form meshwork-like structures. As a consequence, the biofilm became looser and the volume increased accordingly. With the addition of 8 mM R, the EPS were hydrolyzed more evenly. Figure 5E demonstrated that the surface to biovolume ratio of 0.01% Amy + 8 mM R was much higher than the control group as well as 0.01% Amy group, facilitating the accessibility of antimicrobial agents contacting with the cell surface. Brendan revealed that glycoside hydrolase (Sph3_h) could potentiate the treating effects of posaconazole on Aspergillus fumigatus via disrupting the biofilm (Snarr et al., 2017), revealing the analogous mechanism with Figure 5E.

MD simulations predict how each atom in a protein or other molecular substance will move over time, based on experimental structural biology data (Karplus and McCammon, 2002). These simulations can capture a wide variety of important biomolecular processes, including protein folding, ligand binding, conformational changes, and revealing the positions of all the atoms at femtosecond temporal resolution (Hollingsworth and Dror, 2018). Hence, MD simulations could be exploited to determine how a biomolecular system will respond to some perturbation. A number of MD packages, such as CHARMM, AMBER, GROMACS, and LAMMPS could be used to perform biological macromolecular simulations (Lee et al., 2016). Among them, however, GROMACS might be the fastest MD package for its huge codes constantly written by the developers. In addition, the GROMACS analysis facilities for post-processing trajectories are quite extensive, and many other tools could increase a researcher’s productivity, regardless of the simulation package used. Thus, the total time to solution can be minimized by incorporating GROMACS in simulation and analysis (Vermaas et al., 2016). In this study, the molar concentration of R in parts 2.2–2.8 was 8 mM, while the molar concentration of Amy was about 6.4 × 10^–5 mM. Therefore, the molecular amount of R was much higher than Amy. In order to optimize the MD simulation process for energy minimization, R was set decuple than Amy in part 2.9. The MD results showed that 6 R molecules interacted directly with Amy, the other four molecules were dissociated in the force field. PDB 1UA7 is a compound of Amy, containing 422 amino acids, from Bacillus Subtilis complexed with acarbose. The crystal structure of the Amy is divided into three distinct domains (Figure 8A), namely Domain A, Domain B, and Domain C. Domain A (PRO4-ILE100, THR152-LEU352) comprises the typical (α/β)8-Barrels, which is the catalytic domain. Domain B (ASN101-ASN151) is a short loop ring structure extending from Domain A, characterized to hold tightly to a calcium ion, thus being pivotal to the stability of protein. Domain C (SER353-ASP425) comprises a typical antiparallel β-sheet structure. McCorvy et al. (2018) showed that the modified indole-aripiprazole hybrid compounds uncovered the ligand egress of G protein-coupled receptors, indicating the positive response of the receptor to ligand. Dexamethasone privileged only a few poses for the glucocorticoid receptor, providing high rigidity to receptor-ligand complex, thus suitable for recognizing substrates (Alves et al., 2020). Likewise, the RMSD values were reduced and the absolute values of docking scores were significantly improved with the addition of 10 R, implying the potential functional changes of Amy affected by R. Metal-organic frameworks (MOFs) with the shortest inter active site distance (15.6 Å) showed the highest record apparent quantum efficiency in good accordance to biological systems, indicating that a smaller distance leads to higher activity (Gong et al., 2020). In the same vein, greater distances between the catalytic triad in mutant Solanum tuberosum resulted in the prevention of hydrogen bonding which is critical for catalytic activity (Hussain and Chong, 2017). In the present study, similarly, the distances between the catalytic triad were shortened by the binding of the catalytic region to R (Figures 8D,E), which consequently potentiated the catalytic activity of Amy. Flexibility is a key feature of proteins to maintain local changes in conformation (Hollingsworth and Dror, 2018). Figure 8G illustrates the most notable conformational change of Amy with the addition of R, corresponding with the highest peak in RMSF, implying that this random coil from GLY309 to SER320 might reflect the overall stability of Amy.

Comprehensively, based on the results of biofilm biomass, EPS hydrolysis, biofilm morphology, extracellular matrix structure, MD simulations, and molecular docking, 8 mM R could optimally enhance the disassembly effects of 0.01% Amy on A. viscosus biofilm. Still, there exist some limitations. For instance, the effects of Amy + R on the biofilms cultured on the titanium surface should be further investigated. In addition, the feasibility of Amy + R on treating multispecies biofilms should be verified in the next phase of development.

In summary, this study demonstrates that D-arginine may enhance the disassembly effects of alpha-amylase on Actinomyces viscosus biofilm through potentiating the catalytic triad as well as stabilizing the calcium-binding regions, thus providing a novel strategy for the decontamination of biofilm contaminated implant surface.