The study investigated the effect of incorporating l -arginine (Arg) in a glass ionomer cement (GIC) on its mechanical properties and antibacterial potential.
Pre-determined proportions (1%, 2%, and 4% by wt.) of Arg were incorporated in GIC powder; while GIC without Arg served as control. The flexural strength, nanohardness, surface roughness, elemental analysis using SEM-EDX (n = 6) and F/Arg/Ca/Al/Si release in deionized water for 21 days were assessed. The antibacterial potential was evaluated in a multi-species biofilm model with Streptococcus mutans , Streptococcus sanguinis , Streptococcus gordonii , and Lactobacillus acidophilus for 72 h. Real-time qPCR was used to analyse biofilm bacterial concentrations. Propidium monoazide modification of real-time qPCR was performed to quantify viable/dead bacteria. The pH, lactic acid, ADS activity, and H 2 O 2 metabolism were measured. Confocal microscopy was used to investigate the biofilm bacterial live/dead cells, density, and thickness.
There was no difference in flexural strength among the different groups (p > 0.05). No significant difference in nanohardness and surface roughness was observed between 4% Arg + GIC and control (p > 0.05). The 4% Arg + GIC showed significantly higher F/Arg/Al/Si release than the other groups (p < 0.05), reduced total bacterial concentration and growth inhibition of viable S. mutans and S. sanguinis (p < 0.05). Lactic acid formation for 4% Arg + GIC was significantly higher than 1% Arg + GIC (p < 0.05). The spent media pH of 4% Arg + GIC was higher than the other groups (p < 0.05), with proportionately lower ammonia and higher H 2 O 2 released (p < 0.05).
Addition of 4% l -arginine in GIC enhanced its antibacterial activity via a biofilm modulatory effect for microbial homeostasis, with no detrimental effect on its mechanical properties.
Glass ionomer cement (GIC) is widely used in paediatric and geriatric patients as well as for atraumatic restorative treatments in community-based oral health programs. It is well known for its cariostatic properties due to its ability to release fluoride (F). Despite the bioavailability of F to inhibit demineralization and enhance remineralization, secondary caries is still the major cause of failure of GIC restorations [ ]. This can be attributed to the limited effect of F on biofilm control [ ]. It has been shown that F release from GIC is not sufficiently potent to inhibit the growth of cariogenic bacteria [ , ] The presence of bacterial biofilm adjacent to the margins of GIC restoration can lead to its breakdown via microleakage and secondary caries development. Thus, considering the wide clinical applications, chemical adhesive potential, and biocompatibility of GIC; it is imperative that the material be supplemented with either biofilm inhibitory or biofilm modifying agents that prevent the formation of biofilm-mediated caries adjacent to the restoration.
Incorporating biofilm-targeted means in GIC to prevent secondary caries formation have been attempted in the past with most of the approaches being biofilm inhibitory [ , ]. Lately, the use of biofilm modifiers in achieving biofilm homeostasis for caries prevention have been recommended [ ]. Biofilm modifiers enhance the growth of pro-healthy bacteria with subsequent inhibition of pathogenic bacteria. Restoring the diverse biofilm ecology to homeostasis will prevent the caries-associated dysbiosis, averting the detrimental effects associated with the development of pathogenic biofilms adjacent to a restoration. Therefore, incorporating a biofilm modifier at an optimum concentration without deteriorating the physical, chemical, and mechanical properties of GIC seems promising.
Arginine (Arg), a semi-essential amino-acid, is regarded as a prebiotic-based biofilm modifier. The incorporation of the oral prebiotic in fluoridated toothpaste formulations has demonstrated a superior caries-preventive effect compared to the matched controls (including F) [ , ]. Arginine is metabolized by Streptococcus sanguinis, Streptococcus parasanguinis, and Streptococcus gordonii , producing ammonia that neutralizes the acidic environments in the oral cavity [ , ]. This amino acid has been shown to inhibit the growth of Streptococcus mutans and enhance the growth of S. sanguinis in the presence of F [ ]. Furthermore, arginine inhibits pathogenic biofilm matrix and restores biofilm pH homeostasis. It has the potential to modulate the oral microbial community by maintaining healthy oral biofilms [ ]. With the evident biofilm modifier-based caries-preventive effect, the addition of Arg to GIC may impart biofilm homeostatic potential and prevent secondary caries development at the tooth-GIC interface.
In a recent study, Arg has been incorporated into a two-step etch-and-rinse adhesive system based on the premise that Arg would provide a prebiotic-based buffering capacity for secondary caries prevention [ ]. The primer and adhesive comprised of methacrylate and di-methacrylate monomers, ethanol, and photo-initiators with Arg ( l -arginine) at a weight concentration of 5%, 7%, and 10%. The laboratory study concluded that the incorporation of 7% Arg in adhesives exhibited profound antibacterial effects with no alteration of the mechanical properties of the adhesives [ ]. However, no study has so far reported the effect of Arg incorporation in GIC on its mechanical properties and anti-caries potential. Thus, the aim of this study was to investigate the effects of incorporating arginine in GIC on its mechanical properties and antibacterial potential. The null hypotheses tested in the present study were that incorporating Arg in GIC does not affect (1) its mechanical properties and (2) its antibacterial potential.
Materials and methods
GIC formulation and specimen preparation
The experiment and control groups in the study were:
Group 1: 1% Arg in GIC (1% Arg)
Group 2: 2% Arg in GIC (2% Arg)
Group 3: 4% Arg in GIC (4% Arg)
Group 4: positive control — GIC without adding Arg (GIC)
For assessing molecular and elemental release, a negative control group was added:
Group 5: negative control — Arcrylic discs of same dimensions (Acrylic)
The GIC chosen for the present study was GC Fuji II (Fuji II, GC Corporation, Tokyo, Japan). l -arginine powder (A-5006, Sigma Aldrich, St. Louis, USA) at concentrations 1%, 2%, and 4% by wt. were thoroughly vortexed using a vortexor (Vortex-2 Genie, Scientific Industries, Bohemia, USA) with GIC powder (GC Corporation) for 60 s prior to use in a sterilin tube (Bijou, Thermo Scientific, Newport, UK). The powder to liquid ratio of GIC (GC Corporation) was maintained as per manufacturer’s instructions (2.7/1.0 g). The experimental and control disc specimens (9 mm diameter × 4.5 mm thickness) for all the experiments were prepared using the cap of a microcentrifuge tube (graduated microcentrifuge tubes 2.0 mL, (Extra Gene, Taichung, Taiwan) as a sterile mould. The open end of the cap was pressed on polyvinyl strip (No. 697 curved blue 10 mm transparent strip, KerrHawe, California, USA) firmly attached to a glass slide (Cat no. 7105 microscope slides, Sai Brand, Hangzhou, China) after materials were mixed on a mixing paper as per manufacturer’s instructions and introduced to the mould. The material was carefully inserted into the mould using a disinfected agate spatula (GC Fuji Spatula, GC Corporation, Tokyo, Japan) to avoid air bubbles under a sterile hood (SG403AINT SterilGARD III Advance, The Baker Company, Sanford, USA). The cement pressed on glass slide (Sai Brand) was set in place for 10 min. prior to further experimental procedure or assessment. The set cement was carefully removed to prevent any specimen damage. Except for the specimens prepared for the analysis of biaxial flexural strength, all samples were stored in an incubator (Series incubator-Binder, Thomas Scientific, Swedesboro, USA) at 37 °C for 24 h in a relatively humid environment. For biofilm experiments, the specimens were sterilized with UV radiation (SG403AINT SterilGARD III Advance, The Baker Company, Sanford, USA) for 20 min. each side prior to subjecting the discs to bacterial inoculum.
Biaxial flexural strength
The set specimen discs (n = 6) were loaded on a custom-made support ball and the piston tip was positioned at the center of the discs on a computer-controlled universal testing machine (ElectroPuls E3000, Instron, Norwood, United States) with software Bluehill 2 (Bluehill v.2.22, Instron, Norwood, United States). The piston displacement rate was set at 0.5 mm/min and the testing machine auto-recognised the resultant fracture of the discs. The biaxial flexural strength (in MPa) was calculated using the following formula as per previous studies [ , ]:
P — load at facture (N)
d — thickness of the specimens (mm)
v — Poisson’s ratio
A — support ball radius (mm)
B — piston tip radius (mm)
C — specimen radius (mm)
The surface nanohardness (in GPa) of the specimens (n = 6) was determined using atomic force microscopy (Dimension Edge, Bruker, Billerica, USA) with 3 indentations placed at randomly selected points on the disc surface using a diamond indenter (PDNISP, Bruker, Billerica, USA). The scanning mode was set and the indention area was calculated using the software NanoDrive Controller (NanoDrive v.8.03, Bruker, Billerica, USA). The estimated indented areas were used to compute the nanohardness using the following formula:
F — force in N
A — area in mm 2
The surface roughness in μm (Ra) of the specimens (n = 6) was analysed using atomic force microscopy (Bruker) at 3 randomly selected interest areas (20 μm 2 ) to estimate surface scanned height profiles using a scanning probe (OTESPA, Bruker, Billerica, USA). A mean Ra at the combined interest areas was computed to calculate Ra for the specimen using NanoScope Analysis (NanoScope v.1.50, Bruker, Billerica, USA). The topographical assessment was simultaneously performed for quantitative and qualitative evaluation.
Elemental analysis with SEM-EDX
Similar to Ra, 3 randomly selected interest areas were used to determine Ca, Si, Al, and F wt. % in the disc specimens (n = 6) using SEM-EDX (Model 550i, IXRF Systems, Austin, USA) at a standard 500× magnification with SEM (SU1510, Hitachi, Ibaraki, Japan). Along with the inherent elements for GIC (GC Corporation), N wt.% was determined as Arg is a N-containing molecule. Mean wt.% of individual elements were calculated for each experimental group and control to identify any differences in the elemental composition.
Molecular and elemental release
After 24 h storage in an incubator (Thomas Scientific), the specimens (n = 3) were immersed in 5 mL of deionized water (DIW) for 1, 2, 3, 7, 14, and 21 days in an incubator (Thomas Scientific) at 37 °C. Prior to removing the specimens from the medium, the sterilin tubes (Thermo Scientific) were vortexed for 5 s. The experimental medium was then subjected to F, Arg, Ca, Al, and Si release determination. Cumulative release profiles were charted for further data analysis.
The F release was estimated using F-ion selective electrode (Orion 9609BNWP, Thermo Scientific, Newport, UK) subjected to calibration using 0.1, 1, 10, 100, 1000 μg/g F standards (Orion, Thermo Scientific, Newport, UK). The electrode was attached to a bench-top potentiometer (2700 OAKTON, Eutech Instruments, Singapore) with an auto-read mode. The media was subjected to equal volume of TISAB II prior to F analysis. During the measurement process, the solution was continuously stirred using micro-magnetic bars and magnetic stirrer (WiseStir, Daihan Scientific, Seoul, South Korea) at 250 rpm. The electrode was stationed at the centre of the magnetic stirrer with membrane-contact at the fluidic interface. All values were estimated in mV potential and then computed using standards-based regression analysis (R 2 > 0.99) to derive F concentrations at set time-intervals.
The Arg release was determined using o-Pthaldialdehyde derivatization fluorescence-based end-point spectrophotometric method. Arg standards (8-point) from serially diluted 10 μg/g Arg were prepared. A working solution containing o-Pthaldialdehyde (P1378, Sigma Aldrich, St. Louis, USA), absolute ethanol, β-mercaptoethanol, and 0.2 M sodium carbonate (Sigma Aldrich, St. Louis, USA) was prepared. A ratio of 20:1 (working solution: specimen aliquots) was introduced in Costar 96-well black polystyrene plates (Corning Costar, Corning Inc., New York, USA) and read at excitation (OD 340nm )/emission (OD 455nm ) in a UV–vis spectrophotometer (SpectraMax M2, Molecular Devices, San Jose, USA) set for fluorescence end-point evaluation with the software (SoftMax Pro 6 v. 6.33, Molecular Devices, San Jose, USA).
Calcium, aluminium, and silicon release
Around 3 mL of specimen solutions were subjected to Ca, Al, and Si release estimation using high-resolution inductively coupled plasma — optical emission spectroscopy (ICP–OES) (Spectro Arcos, Ametek, Kleve, Germany) as per our previous study [ ]. Standard solutions for each element were prepared at concentrations 0, 1.56, 3.13, 6.25, 12.50, 25, 50, and 1000 μg/g (TraceCERT®, Sigma-Aldrich, St. Louis, USA). The spectrum was calibrated using negative and positive controls. The measurements in μg/g were made using Smart Analyser Vision, 2014 v. 6.01.0943 (Spectro Analytical Instruments, GmBH, Kleve, Germany) utilized for spectrum calibration.
Multi-species biofilm model
S. mutans ATCC 35668, S. sanguinis ATCC 10556, S. gordonii ATCC 35105, and Lactobacillus acidophilus ATCC 9224 (American Type Culture Collection ATCC, Manassas, USA) were cultured in BHI (53286, Sigma-Aldrich, St. Louis, USA). The bacterial cells for each strain were adjusted to the concentration of 10 7 CFU/mL using McFarland spectrophotometric method at OD 600nm (Spectrophotometer, DU 730 Life Science, Beckman Coulter, California, USA). Multi-species inoculum was prepared in the ratio of 1:1:1:1. The sterilized specimens were introduced in a 24-well plate and inoculated using the prepared inoculum medium for 72 h in an anaerobic chamber (85% N 2 , 10% H 2 , 5% CO 2 ; 37 °C). The biofilms containing specimens were then removed and dip-washed in PBS prior to further investigations.
DNA isolation and real-time qPCR
The total (live + dead) bacterial concentrations for each strain were determined using molecular-based real-time qPCR method as per our previous study [ ]. The dip-washed biofilms were suspended in 0.9% NaCl and vortexed for 60 s. Then, the specimens were carefully removed and the suspended biofilms were centrifuged at 14,000 × g for 10 min. The supernatant was discarded and the pellet was suspended in 20 mM Tris-HCl (pH-8.0); 2 mM EDTA; 1.2% Triton buffer (Sigma-Aldrich, St. Louis, USA). The biofilm bacterial cells were lysed with 20 mg/mL lysozyme (Sigma-Aldrich, St. Louis, USA) with an overnight water-bath incubation at 37 °C. The DNA isolation was done using QIAmp DNA isolation kit (QIAmp, Qiagen, Hilden Germany) as per manufacturer’s instructions. The positive controls for DNA quantification (reference ATCC strains) underwent the same steps as the bacterial pellet with 5 different concentrations/strain to obtain the regression-based standard curve.
The primers used in the present study are shown in Table 1 . The SYBR green master PCR mix (FAST SYBR Green Master PCR mix, Thermo Fisher Scientific, Waltham, USA), isolated DNA, primers (Life technologies, Thermo Fisher Scientific, Hong Kong), and MilliQ water were introduced in a 96-well optical PCR reaction plate (MicroAmp®, Thermo Fisher Scientific, Waltham, USA) with negative (MilliQ water) and positive controls. The real-time qPCR reaction was initiated using Step One Plus (Real-time PCR System, Applied Biosystems, Waltham, USA) with cycle conditions: 50 °C/2 min.; 95 °C/10 min.; 50 cycles of 95 °C/15 s and 58 °C/1 min. The bacterial concentrations for each strain per biofilm were estimated based on the derived standard curve.
|S. mutans||F: 5′ – GCCTACAGCTCAGAGATGCTATTCT – 3′|
|R: 5′ – GCCATACACCACTCATGAATTGA – 3′|
|S. sanguinis||F: 5′ – CAAAATTGTTGCAAATCCAAAGG – 3′|
|R: 5′ – GCTATCGCTCCCTGTCTTTGA – 3′|
|S. gordonii||F: 5′ – CGGATGATGCTAATCAAGTGACC – 3′|
|R: 5′ – GTTAGCTGTTGGATTGGTTGCC – 3′|
|L. acidophilus||F: 5′ – GAC TGC AAA GTG GTA GCG TAA GC – 3′|
|R: 5′ – CCG GCC TAC TCA GGA TTC TG – 3′|
Propidium monoazide (PMA) modification of real-time qPCR
Prior to all the steps for DNA isolation and real-time qPCR (as per 2.8.) as utilized for PMA modified real-time qPCR, the suspended biofilms in 0.9% NaCl were equally divided into two fractions. A fraction was subjected to PMA stain (PMAXX 40069, BOTIUM, Fremont, USA) and ice incubated on an orbital shaker (60 rpm, room temperature) for 10 min in a biosafety cabinet (The Baker Company). Then, the PMA stained biofilm suspensions were held for photoactivation of PMA using LED photolysis unit (PMA-Lite™, Biotium, Fremont, USA) for 5 min. Following PMA staining, care was taken to avoid light exposure to the suspensions. The biofilm suspensions were then centrifuged at 9880 rpm for 10 min to receive the bacterial pellet for DNA isolation followed by real-time qPCR (Applied Biosystems). The quantified viable/dead cells were calculated based on the total bacterial concentration estimated in unstained sub-section serving as a control.
pH, ADS activity, lactic acid, and H 2 O 2 measurements
The spent media in the 24-well microplate after 72 h was subjected to pH, ADS activity, lactic acid and H 2 O 2 measurements. The pH of the spent media was measured using a sensitive pH electrode (Oakton WD-35805-04, Eutech Instruments, Singapore) attached to a bench-top potentiometer (Eutech Instruments). An auto-calibrated pH standard curve was established using pH –4.01, 7.00, and 10.01 external standards (Orion, Thermo Scientific, Waltham, USA). The potentiometer measured the media pH using the auto-read measurement mode indicating the measured pH stability. All measurements were made in triplicate.
Lactic acid was measured as per a previous study in g/L [ ]. Briefly, the spent media was centrifuged at 10,000 rpm for 5 min. The supernatant aliquot was suspended in 0.2% FeCl 2 solution (Sigma-Aldrich, St. Louis, USA). A standard series of lactic acid (Sigma-Aldrich, St. Louis, USA) from 89 g/L lactic acid was prepared within the suggested calibration range. All lactic acid measurements were made spectrophotometrically at OD 390nm in triplicate (Molecular Devices).
ADS activity was determined based on the amount of ammonia produced in the spent media. Equal volume of test media and Nessler’s reagent (Sigma-Aldrich, St. Louis, USA) subjected to 8-point NH 3 standards (from ammonium sulphate)-based regression estimated the NH 3 concentration in μg/g using a micro-plate. The ammonia was determined spectrophotometrically at OD 395nm (Molecular Devices).
Hydrogen peroxide was estimated using H 2 O 2 assay kit (Abcam, Cambridge, USA). The spent media was centrifuged at 1000 × g for 30 min at 4 °C to receive the supernatant. The supernatant was subjected to deproteination prior to estimating the H 2 O 2 concentration. Briefly, 4M HClO 4 (ice cold) was added to the supernatant to finally receive a 1M HClO 4 concentration solution, vortexed and incubated on ice for 5 min. The acid diluted supernatant was centrifuged at 13,000 × g for 2 min at 4 °C. The supernatant was then supplemented with ice cold 2M KOH (40% of supernatant) and vortexed to receive the neutralised samples. Using pH colorimetric assay (Hydrion Buffer Color Key Preservative, Micro Essential Labs, Brooklyn, USA), the pH of the supernatant was adjusted to 6.5–8 with 2M KOH (ice cold). After establishing the pH, the neutralised mix was centrifuged at 13,000 × g for 2 min at 4 °C. The supernatant was collected for estimating H 2 O 2 concentration. The sensitive H 2 O 2 assay fluorometric method utilized a reaction mix of assay buffer, oxired probe, and horse radish peroxidase (Abcam). Equal aliquots of deproteinized/neutralized supernatant and reaction mix were suspended in the 96-well black polystyrene plate (Costar) with 0.1–0.5 nmol H 2 O 2 standards. The plate was measured using a spectrophotometer (Molecular Devices) at Ex/Em–OD 535nm /OD 587nm . The OD values were extrapolated based on the established standard curve and then derived in μM H 2 O 2 . Finally, the determined ammonia and H 2 O 2 concentrations were derived in μg/g to establish proportional distribution of the biochemical components in the 72 h spent media.
Confocal laser scanning microscopy
Biofilms bacterial live/dead cells were characterized using LIVE/DEAD® BacLight ™ Bacterial Viability Kit (Molecular Probes, Grand Island, USA). The dip-washed biofilms were stained with SYTO9/Propidium Iodide and scanned with two-photon laser scanning microscope (FLUOVIEW FV1000, Olympus, Center Valley, USA). The scanning was done at 3 randomly selected distinct interest areas within the confines of the biofilm at 100×. A 3-D scanned biofilm stack was re-constructed to measure the biofilm thickness in μm. The biofilm bacterial vitality was determined qualitatively by identifying appropriate red (dead) and green (live) bacterial cells while determining the bacterial density in the biofilm.
The data was entered in MS Excel software (Microsoft, Redmond, USA) and further subjected to statistical analysis using SPSS v. 25 (IBM statistics, New York, USA). The parametric 1-way ANOVA with Tukey’s HSD post-hoc test was used to analyse the data for surface roughness and nanohardness, biaxial flexural strength, F/Arg/Ca/Al/Si release (independent, cumulative, and integrated), biofilm thickness, total bacterial concentration with real-time qPCR, pH, ADS activity, H 2 O 2 concentration, and generated lactic acid.
Kruskal–Wallis 1-way ANOVA with Dunn’s post-hoc test analysed the data for SEM-EDX determining concentrations of N, F, Al, Si, Ca in different experimental and control groups.
The integrated mean F, Arg, Al, and Si release were subjected to further analysis with Pearson correlation coefficient test to identify bivariate relations between the estimated elements/molecules. Similarly, correlations were established between the mean integrated Arg release and the determined spent media pH.
The determined proportional ammonia and H 2 O 2 concentrations, bacterial concentrations with real-time qPCR, viable and dead bacterial concentrations with PMA modified real-time qPCR were analysed using 2-way ANOVA with LSD test. Bivariate correlations were determined between the Arg concentrations and live/dead (total) bacterial concentrations. The statistical significance for all the tests was set at α = 0.05 whereby the experimental variables outcome were replicate measurements at disparate time-points.
Mechanical properties and surface roughness
The mechanical properties and surface roughness of the experimental and control groups are shown in Fig. 1 . No significant difference in the biaxial flexural strength could be identified among the experimental and control groups (p > 0.05) ( Fig. 1 (A)). No significant difference in the nanohardness and surface roughness was seen between the 4% Arg and control GIC groups (p > 0.05). However, the nanohardness and surface roughness of the 1% and 2% Arg groups were significantly lower than the control GIC group ( Fig. 1 (B–D)).