A Combination of Zinc and Arginine Disrupt the Mechanical Integrity of Dental Biofilms

ABSTRACT Mechanical cleaning remains the standard of care for maintaining oral hygiene. However, mechanical cleaning is often augmented with active therapeutics that further promote oral health. A dentifrice, consisting of the “Dual Zinc plus Arginine” (DZA) technology, was found to be effective at controlling bacteria using in vitro laboratory studies, translating to clinical efficacy to deliver plaque and gingivitis reduction benefits. Here, we used biophysical analyses and confocal laser scanning microscopy to understand how a DZA dentifrice impacted the mechanical properties of dental plaque biofilms and determine if changes to biofilm rheology enhanced the removal of dental plaque. Using both uniaxial mechanical indentation and an adapted rotating-disc rheometry assay, it was found that DZA treatment compromised biofilm mechanical integrity, resulting in the biofilm being more susceptible to removal by shear forces compared to treatment with either arginine or zinc alone. Confocal laser scanning microscopy revealed that DZA treatment reduced the amount of extracellular polymeric slime within the biofilm, likely accounting for the reduced mechanical properties. We propose a model where arginine facilitates the entry of zinc into the biofilm, resulting in additive effects of the two activities toward dental plaque biofilms. Together, our results support the use of a dentifrice containing Dual Zinc plus Arginine as part of daily oral hygiene regimens. IMPORTANCE Mechanical removal of dental plaque is augmented with therapeutic compounds to promote oral health. A dentifrice containing the ingredients zinc and arginine has shown efficacy at reducing dental plaque both in vitro and in vivo. However, how these active compounds interact together to facilitate dental plaque removal is unclear. Here, we used a combination of biophysical analyses and microscopy to demonstrate that combined treatment with zinc and arginine targets the matrix of dental plaque biofilms, which destabilized the mechanical integrity of these microbial communities, making them more susceptible to removal by shear forces.


RESULTS
Streptococcus gordonii biofilms treated with Dual Zinc plus Arginine were more susceptible to mechanical removal. An understanding of biofilm mechanics is important to the development and application of therapeutics because these properties often define how a biofilm responds to chemical and mechanical forms of eradication. Furthermore, changing the biofilm mechanics is associated with improved penetration and, therefore, enhanced action of therapeutic compounds (31,32). We, therefore, sought to understand how the DZA technology altered the mechanics of biofilms relevant to oral health. We recently adapted a novel rotating-disc rheology assay to analyze and quantify the detachment of arginine-treated Streptococcus gordonii biofilms (17). S. gordonii is a natural commensal of saliva-plaque biofilms and supportive of early colonizers of dental plaque (6,(33)(34)(35). We used this assay to determine how treatment with DZA impacted S. gordonii biofilm mechanics and removal when exposed to shear stress.
S. gordonii biofilms were grown on coupons for 5 days, after which they were treated with either PBS (untreated control), Dual Zinc (0.96% zinc ions; zinc oxide and zinc citrate), or DZA (0.96% zinc ions and 1.5% L-arginine) for 2 min. Biofilm-coated coupons were then attached to the rheometer and spun at increasing velocities from 0.1 to 300 rad/s over 360 s and immersed in water. The resulting torque, a measurement of resistance to rotation of the coupon, was measured across the velocity range. This analysis revealed that treated S. gordonii biofilms displayed a reduced torque compared to untreated biofilms ( Fig. 1A to C). This indicated that coupons coated with treated S. gordonii biofilms had less resistance to rotation compared to untreated biofilms. Furthermore, depicting these data as the mean of individual replicates revealed that the torque of S. gordonii biofilms treated with DZA was lower across the assayed velocity range compared to both the untreated and zinc-treated biofilms (Fig. 1D).
We previously observed that reductions in torque during the test correlated with the detachment of biofilm aggregates from the coupon surface (17). These biofilm detachment events were similarly observed here (Fig. 1). To depict these events, a baseline correction was performed by which the data were linearly transformed and presented more clearly as the change in slope between measurements (i.e., an approximation of the first derivative), which enhanced the changes in torque as observed by the negative slope values (Fig. 2). This revealed that treatment of S. gordonii biofilms led to reductions in torque at lower velocity ranges compared to untreated biofilms (Fig. 2). Moreover, changes in torque at lower velocities were observed for biofilms treated with DZA, compared to untreated and zinc treated biofilms ( Fig. 2; green bracket).
To quantify these differences the biofilm momentum coefficient was determined, which is an indication of the biofilm-induced drag on rotation (17,36,37). Although there were no statistical differences between the biofilm momentum coefficient between treated and untreated biofilms, there was a general trend that both zinc-and DZA-treated S. gordonii biofilms had a reduced coefficient compared to untreated biofilms. This was more pronounced for DZA-treated biofilms ( Fig. 3A; biofilm momentum coefficient mean 6 SD; untreated 0.063 6 0.021; Zn 0.053 6 0.023; DZA 0.037 6 0.023). This suggested that less drag was experienced by coupons coated with DZA-treated biofilms, which is associated with a less adhered biofilm (17,36,37). To investigate this further, the area under the curve (AUC) of the torque-angular velocity curves were measured, which is an indication of the amount of energy required for the coupon to rotate. This analysis revealed that only DZA-treated biofilms had a significantly reduced AUC compared to untreated FIG 1 S. gordonii biofilms treated with Dual Zinc plus Arginine display reduced torque when analyzed by adapted rotating-disc rheology. S. gordonii biofilms were grown for 5 days and were treated with either (A) PBS (untreated), (B) 0.96% zinc (Zn), or (C) DZA for 2 min. Biofilms were analyzed by adapted rotating-disc rheology. Data are presented as torque-angular velocity curves of individual replicate biofilms, separated per treatment group for clarity. (D) Data from (A to C) presented as the mean 6 95% confidence interval of combined replicates for direct comparison across treatments. Three biological replicates were performed each with duplicate biofilms. biofilms (Fig. 3B), corroborating the trends observed from the biofilm momentum coefficient analysis (Fig. 3A). Together, these data indicated that coupons coated with DZAtreated biofilms experienced less drag compared to untreated and zinc treated biofilms, suggestive of a less adherent and mechanically weaker biofilm.
To determine if these observations correlated with differences in biofilm-aggregate detachment, the angular velocity where the first reduction in torque occurred was converted to a shear stress, indicative of the initiation or critical biofilm detachment shear stress (Fig. 3C). This revealed that reductions in torque, which correlated with biofilmaggregate detachment, occurred at significantly lower shear stresses for treated S. gordonii biofilms compared to untreated biofilms (Fig. 3C). Furthermore, DZA-treated biofilms had significantly lower critical biofilm detachment shear stresses compared to zinc treated biofilms (Fig. 3C). This indicated that DZA-treated biofilms detached from the coupon surfaces at lower shear stresses compared to untreated and zinc treated biofilms, suggesting greater susceptibility to removal by external shear forces.
Combined treatment with Dual Zinc plus Arginine destabilized the mechanical integrity of saliva-plaque biofilms. Having demonstrated that treatment with DZA led to the mechanical destabilization of S. gordonii mono-species biofilms ( Fig. 1 to 3), we wanted to determine the effects of DZA treatment using a biofilm model that more closely mimics in vivo conditions.
To assess this, saliva-plaque biofilms were grown on hydroxyl-appetite (HA) discs for 5 days (38,39). Biofilms were then treated with either solution of PBS (untreated control), 1.5% arginine, Dual Zinc, or DZA for 2 min. The biofilm mechanical properties were then assessed using uniaxial mechanical indentation, where biofilms were compressed at a defined rate and the resistance force measured. The biofilm thickness was determined by an increase of force in resistance to compression. The treatment did not change the biofilm thickness (Fig. 4A). The differences in the resulting stress-strain profiles revealed that all three treatments  altered the biofilm mechanical properties relative to the untreated control ( Fig. 4B). To therefore quantify the differences in the biofilm mechanical properties, the Young's modulus, a measure of the ability of the biofilm to resist compression and a measure of biofilm stiffness (31,40), was determined. This revealed that biofilms treated with either arginine, zinc, or DZA had a significantly lower Young's modulus compared to the untreated control (Fig. 4C). This indicated that treatment with either solution destabilized the mechanical integrity of the saliva-plaque biofilm. While there was no statistically significant difference (P > 0.05) between either treatment, there was a trend where biofilms treated with arginine or zinc alone or DZA showed incremental decreases in the Young's modulus. Biofilms treated with DZA displayed the lowest Young's modulus (Young's modulus mean 6 SD; untreated 27.28 6 8.68; arginine 15.38 6 7.97; zinc 11.14 6 5.02; DZA 9.29 6 2.87). This suggested that treatment with DZA had the most impact on destabilizing the mechanical properties of saliva-plaque biofilms.
We were also interested in determining if the ability of DZA to destabilize the biofilm mechanical properties was maintained when developed into a dentifrice paste. Salivaplaque biofilms were grown for 5 days and treated with either DZA paste, a commercially available stannous fluoride toothpaste, or PBS and analyzed by mechanical indentation as above. This revealed that biofilms treated with DZA dentifrice paste had a significantly lower Young's modulus compared to untreated biofilm ( Fig. S1 in Supplemental File 1), which was similar to what we observed for treatment with the test solutions (Fig. 4C). In contrast, biofilms treated with the stannous fluoride control paste had a significantly higher Young's modulus compared to both untreated biofilms and biofilms treated with DZA ( Fig. S1 in Supplemental File 1). This demonstrated that both the action of DZA on biofilms was maintained in a toothpaste formula and that the DZA technology may be more effective at reducing the mechanical integrity of saliva-plaque biofilms compared to other commercially available antibacterial dentifrices.
Combined treatment with Dual Zinc plus Arginine degraded the EPS of salivaplaque biofilms. Biofilm mechanics is a property associated with the collective behavior of the biofilm that is largely imparted by the extracellular polymeric slime (EPS) (31). Having observed that treatment with either compound did not result in changes to the biofilm thickness (Fig. 4B), we therefore, wanted to determine if treatment with DZA exhibited an impact on biofilm EPS.
To assess this, 5-day saliva-plaque biofilms were treated with either PBS (untreated control), Dual Zinc, or DZA for 2 min. Treatment with arginine alone was omitted from this analysis because it showed an overall higher Young's modulus compared to the other two treatments (Fig. 4C). Biofilms were then stained with Syto 9 to label the bacterial cells and calcofluor white to label the EPS and imaged using confocal laser scanning microscopy (CLSM). This revealed that treatment with either zinc or DZA appeared to reduce the amount of calcofluor white-labeled EPS within the biofilm (Fig. 5A).
To quantify these changes, both Syto 9 and calcofluor white staining was analyzed using COMSTAT (41,42). Consistent with the biofilm thickness analysis from the uniaxial indentation measurements (Fig. 4B), neither treatment significantly altered the biomass of cells compared to the untreated control (Fig. 5B). However, calcofluor white-stained EPS revealed that treatment with either zinc or DZA showed a significant reduction in the volume of EPS within biofilms (Fig. 5C). Furthermore, the roughness coefficient, a measurement of the variability in height and heterogeneity (41,42), of calcofluor white-labeled EPS of treated biofilms was significantly higher than untreated controls (Fig. 5D). This indicated that treatment with either zinc or DZA led to an overall reduction in calcofluor white-labeled EPS and changed the distribution of EPS from a more homogenous layer in untreated biofilms to a more heterogeneous punctate distribution in treated biofilms. Again, there was a trend that calcofluor white-labeled EPS of DZA-treated biofilms had a higher roughness coefficient compared to treatment with zinc alone (Fig. 5D). This suggested that DZA treatment led to a more uneven, heterogenous EPS distribution within the biofilm compared to treatment with zinc. However, this was not statistically significant.

DISCUSSION
Treating biofilms with DZA is effective at reducing dental plaque biofilms in vitro and in vivo (5,16,29,30). However, it is unclear how DZA impacts the structure and mechanical properties of dental plaque biofilms. Here, we demonstrated that DZA destabilized the mechanical integrity of both S. gordonii mono-species biofilms ( Fig. 1 to 3) and dental plaque biofilms ( Fig. 4 and Fig. S1 in Supplemental File 1) by targeting the EPS (Fig. 5). This correlated with biofilms being more susceptible to removal by shear forces (Fig. 1 to 3). Furthermore, we observed that these effects of DZA appeared to be increased compared to either zinc or arginine treatment alone. This suggests that there are synergistic or additive interactions between zinc and arginine that mediate the disruption of dental plaque biofilms.
We propose the following model for potential interactions between arginine and zinc (Fig. 6). There is evidence to suggest that exogenous arginine can penetrate bacterial biofilms (43) and disrupt cell-cell and EPS interactions within the biofilm (10,15). We, therefore, propose that arginine can penetrate the dental plaque biofilm, targeting interactions within the EPS and destabilizing the mechanical structure of the biofilm (Fig. 6). This is consistent with the reduced Young's modulus observed for arginine-treated saliva-plaque biofilms compared to untreated biofilms (Fig. 4C). Furthermore, we previously reported that arginine-treated S. gordonii biofilms are more susceptible to removal by shear forces, due to the weakened Arginine and Zinc Disrupt Dental Biofilms Microbiology Spectrum mechanical structure (Fig. 6) (17). Here, we also demonstrated that treatment with zinc alone could weaken the mechanical structure of dental plaque biofilms ( Fig. 4 and Fig. S1 in Supplemental File 1) by targeting the biofilm EPS (Fig. 5). Similarly, zinc has been shown to inhibit surface attachment and EPS production of Streptococcus mutans (44). However, the antimicrobial effects of zinc are reduced in biofilms compared to planktonic cultures, and it has been suggested that zinc is unable to penetrate the biofilm (19,23,26). We, therefore, propose that zinc can target the dental plaque biofilm EPS possibly through cationic interactions. However, this action is limited to the periphery of the biofilm (Fig. 6). We predict that zinc treatment weakens the mechanical structure of the biofilm to a similar extent as arginine treatment, which is consistent with the Young's modulus of zinc-or arginine-treated saliva-plaque biofilms being similar (Fig. 4C). For treatment with DZA, we predict that arginine first penetrates and structurally weakens the biofilm through disruption of the biofilm EPS. Loosening the biofilm structure then permits zinc to gain entry deeper into the biofilm, extending the effects of zinc to deeper regions of the biofilm (Fig. 6). This results in further weakening of the biofilm mechanical structure and increased susceptibility of the biofilm to removal by shear forces, compared to treatment with either arginine or zinc alone. Supporting this model, matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry (MS) analysis revealed that saliva-plaque biofilms treated with a DZA dentifrice had increased penetration and retention of zinc compared to treatment with Dual Zinc alone (16). This model currently only takes into consideration the physical effects over the short time scales (2 min) analyzed here.
To determine if the mechanical destabilization of dental plaque biofilms was due to synergistic or additive interactions between zinc and arginine, we applied a modified fractional inhibitory concentration (FIC) index, referred to here as the interaction index. The FIC index is traditionally used to determine whether the interactions between antimicrobials are synergistic, additive, or antagonistic when used in combination (45,46). By substituting the MICs for the Young's modulus, according to equations 2 to 4, we obtained an interaction index of 1.87 6 0.95 (mean 6 SD) for DZA, suggestive of additive interactions (45,46). Similarly, when the Young's modulus of arginine-treated and zinc-treated saliva-plaque biofilms are subtracted, we obtained a value (4.95 6 3.30) more similar to the Young's modulus of DZA-treated biofilms (9.29 6 2.87) compared to when moduli were divided (1.51 6 0.35). This further supported that the interactions between zinc and arginine were additive. This suggests that across the short time scales analyzed here (2 min), both arginine and zinc have a similar target, presumably EPS (Fig. 5), that is responsible for the mechanical destabilization of the biofilm. However, across longer time scales with repeat treatments, it is likely that the antimicrobial effects of zinc and the metabolic effects of arginine will influence these interactions, potentially shifting them to being more synergic. Therefore, future studies that focus on these interactions, across longer time scales, are warranted. Furthermore, this study focused on the maintenance of oral health hygiene, rather than the removal of pathogenic oral biofilms. Therefore, biofilms formed by oral commensals (S. gordonii) and saliva-plaque biofilms grown from healthy donors were only examined here. Future studies that focus on the cariogenic oral biofilm pathogen S. mutans grown in the presence of sucrose are also warranted to determine if these interactions between zinc and arginine are maintained in the treatment of a cariogenic biofilm, which include a sticky glucan EPS matrix and acidic environment.
Importantly, long-term (6 months) clinical trials of a dentifrice containing DZA did not report any adverse side effects among the 100 participants (30). Similarly, long-term studies (>1 year) of zinc-supplemented dentifrices reported no significant changes in oral microbial ecology or antimicrobial-resistant bacteria (47,48). Our results and proposed model suggest that targeting dental plaque biofilm mechanics is a strategy to enhance the action of actives against bacteria within dentifrices. Our results are also supported by similar observations, where arginine increased the susceptibility of saliva-plaque biofilms to cetylpyridinium chloride (15), Streptococcus pyogenes biofilms to a penicillin (49), and Pseudomonas aeruginosa biofilms to tobramycin and ciprofloxacin (43). Our results, therefore, support the growing evidence that arginine has the potential to be used as an antimicrobial adjuvant. Finally, zinc oxide inhibits the biofilm formation of P. aeruginosa, Chromobacterium violaceum (50), and Candida albicans (51). This suggests that DZA has the potential to be used across diverse clinical applications for the treatment of biofilm-associated infections, in addition to the maintenance of oral health and hygiene.

MATERIALS AND METHODS
Preparation of treatment formulations. Test solutions consisted of 1.5% arginine, Dual Zinc (0.96% total zinc ions; zinc citrate and zinc oxide), or Dual Zinc plus Arginine (0.96% zinc ions; zinc oxide and zinc citrate, and 1.5% L-arginine). Solutions were adjusted to a final pH of 7.0. In toothpaste-treated biofilm experiments, the biofilms were treated with either commercially available stannous fluoride 0.0454% based active agent stabilized with sodium hexametaphosphate (Procter and Gamble, UK) or Colgate Total containing Dual Zinc plus Arginine and (Colgate-Palmolive, UK).
Saliva-plaque inoculum collection. Saliva plaque was collected from healthy individuals according to protocols approved by the Ohio State University (OSU) Institutional Review Board (IRB) (number 2017H0016, 03/22/2017) (39). Collected samples were pooled and separated into 1 mL aliquots supplemented with 30% glycerol. Samples were stored at 280°C before experimentation.
Biofilm growth. (i) S. gordonii biofilms. S. gordonii wild-type strain DL1 was used in this study. Biofilms were grown as previously described (17). Briefly, sterile 40 mm diameter coupons were submerged in 40 mL brain heart infusion media, supplemented with 0.5% sucrose, and inoculated with 400 mL of overnight S. gordonii culture. Coupons were incubated in a humidified chamber at 37°C with 5% CO 2 and on an orbital shaker at 150 rpm. The growth medium was replaced every 24h. Biofilms were grown for 5 days.
(ii) In vitro saliva-plaque biofilms. Sterile hydroxyl-appetite (HA) discs were transferred to a 24-well plate, one disc per well. Next, 900 mL of McBain media (2g/L peptone, 2g/L Trypticase peptone, 1g/L yeast extract, 0.35g/L NaCl, 0.2g/L KCl, 0.2g/L CaCl 2 , 2.5g/L mucin, 50 mM PIPES buffer, 5 mg/mL hemin, and 1 mg/mL vitamin K) was added to each well and inoculated will 100 mL of pooled saliva-plaque. The plate was incubated in a humidified chamber at 37°C with 5% CO 2 on an orbital shaker at 150 rpm. Every 24h the growth medium was replaced with 1 mL of fresh McBain media. Biofilms were grown for 5 days. This model is supportive of the growth of early colonizer aerobes and late colonizer anaerobes within the saliva-plaque biofilm (38,39).
Biofilm treatment. (i) Treatment with solutions. After 5 days of growth, S. gordonii biofilm-coated coupons were transferred to a dish containing 60 mL of either PBS (untreated control), Dual Zinc (0.96% zinc ions), or DZA. Biofilms were treated for 2 min at 37°C with 5% CO 2 and shaking at 150 rpm. Biofilms were then washed twice in PBS before being transferred into 60 mL PBS until analysis.
Saliva-plaque biofilm-coated hydroxyl-appetite (HA) discs were treated as above except biofilms were transferred to a new 24-well plate, with each well containing 1 mL of either PBS (untreated control), 1.5% arginine, Dual Zinc, or DZA. After treatment biofilms were washed twice in PBS, before being transferred into 1 mL PBS until analysis.
(ii) Treatment with pastes. One gram of DZA paste was dissolved in 4 mL sterile water (1:5 vol/vol dilution). One gram of commercially available stannous fluoride toothpaste was also dissolved in 4 mL of sterile water and used as a control paste treatment. The foam was allowed to settle before 1 mL of paste suspension was transferred to wells of 24-well plate. HA discs coated with 5-day saliva-plaque biofilms were transferred into wells containing paste suspension or sterile water (untreated control). Biofilms were treated for 2 min at 37°C with 5% CO 2 and shaking at 150 rpm. Biofilms were then washed twice in PBS, before being transferred into 1 mL PBS until analysis.
Rheometry analysis. (i) Adapted rotating-disc rheology. S. gordonii biofilms were analyzed using an adapted rotating-disc rheology assay as previously described (17). Briefly, biofilm-coated coupons were attached to the shaft of a Discovery Hybrid Rheometer-2 (DHR-2) (TA Instruments). Biofilms were immersed in a 15 Â 15 cm square clear acrylic container filled with 2.8 L reverse osmosis water with the gap distance set to 3.5 cm. Immersed coupons were spun at an angular velocity (v ) range of 0.1 to 300 rad/s, incrementing the speed across 360 s. For each biofilm treatment (untreated, Dual Zinc, or DZA), 3 biological replicates were performed, each with duplicate biofilms (a total of 6 biofilms analyzed per treatment group).
The torque-angular velocity curves were linearized and transformed to more clearly depict the data as previously described (17). The fouling coefficient and conversion of torque to shear stress were calculated as previously described (17). The area under the curve (AUC) of the torque-angular velocity curves were determined using the analysis function within GraphPad Prism v8.
(ii) Uniaxial mechanical indentation. Biofilms were analyzed on a DHR-2 with a Peltier plate connected to a heat exchanger (TA Instruments). Biofilms were analyzed and submerged in PBS. Indentation measurements were performed at 25°C using the 8 mm Smart Swap geometry with an approach rate of 1 mm/s and a termination detection at 15 N. Contact with the biofilm was determined where the force began to steadily increase. Young's modulus (E) was calculated using the force-displacement relationship previously described (40).
The slope is of the force-displacement curve (N/m), r is the radius of the probe (r = 0.004m), and v is the assumed Poisson's ratio of a biofilm (v = 0.5) (52). The slope of the force-displacement curve was measured at the region corresponding to 0 to 20% strain. Four biological replicates, each with three technical replicates, were performed for each treatment for both solutions and paste suspensions (a total of 12 biofilms analyzed for each treatment group).
To determine if the combined actions of DZA to the biofilm mechanical properties were synergistic or additive, we used the fractional inhibitory concentration (FIC) index, referred to here as the interaction index (IAI), which has traditionally been used to determine the interactions between multiple antimicrobials (45,46). However, we substituted the MIC for the Young's modulus, according to equation 2 to 4.

IAI
Here, IAI # 0.5 was indicative of synergism, IAI > 0.5 to 4 was indicative of additive, and IAI $ 4 was indicative of antagonism. Confocal laser scanning microscopy. For confocal laser scanning microscopy (CLSM), treated 5-day saliva-plaque biofilms were transferred to 1 mL 4% paraformaldehyde and fixed at room temperature for 1 h. Biofilms were then transferred to 1 mL PBS supplemented with 5 mM Syto-9 and 10 mg/mL calcofluor white and incubated for 1 h at room temperature. Biofilms were washed twice in PBS and transferred to fresh PBS until imaging. Biofilms were imaged on an Olympus FV1000 Multiphoton microscope, in single photon mode, fitted with a 25Â water immersion objective. Images were processed using the Imaris software and analyzed using COMSTAT-2 (41,42). For each biofilm treatment (untreated, Dual Zinc, or DZA), 3 biological replicates were imaged, with 4 images taken per biofilm.
Statistical analysis. Statistical analysis was performed using a one-way analysis of variance ANOVA with Tukey's post hoc test. Analyses were performed using GraphPad Prism v8 (GraphPad Software). Statistical significance was determined using P , 0.05.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 0.1 MB.