Hydrogel encapsulated biofilm inhibitors abrogate the cariogenic activity of Streptococcus mutans

Abstract

We designed and synthesized analogs of a previously identified biofilm inhibitor, IIIC5 to improve solubility, retain inhibitory activities, and to facilitate encapsulation into pH-responsive hydrogel microparticles. The optimized lead compound, HA5 showed improved solubility of 120.09 μg/mL, inhibited S. mutans biofilm with an IC₅₀ value of 6.42 μM and did not affect the growth of oral commensal species up to a 15-fold higher concentration. The co-crystal structure of HA5 with GtfB catalytic domain determined at 2.5 Å resolution revealed its active site interactions. The ability of HA5 to inhibit S. mutans Gtfs and to reduce glucan production has been demonstrated. The hydrogel encapsulated biofilm inhibitor (HEBI), generated by encapsulating HA5 in hydrogel, selectively inhibited S. mutans biofilms like HA5. Treatment of S. mutans infected rats with HA5 or HEBI resulted in a significant reduction in buccal, sulcal, and proximal dental caries compared to untreated, infected rats.

INTRODUCTION.

Most tooth and gum related diseases are associated with bacterial infections. Among these, dental caries (tooth decay) is a ubiquitous disease that affects much of the human population. Dental caries is a multifactorial disease that causes localized destruction of susceptible dental tissues¹. Dental caries is identified as the most prevalent disease worldwide in a recent Lancet study of global burden of 328 major diseases². Despite its general classification as a ‘life-style-related’ disease, dental caries poses a significant challenge as it results in tooth loss, infection, and in some cases, even death by sepsis^{3, 4}.

Current treatments for this disease have severe limitations. The conventional oral hygiene practices such as brushing or mouthwashes are not highly effective due to the rapid re-colonization of the bacteria⁵. Fluoride sealants and varnishes are commonly used to prevent dental caries in children⁶. While there is a general consensus on the safety of fluoride treatments⁷, their high fluoride content (1–5 %) and potential neurotoxic effects are a concern⁸. The antimicrobial agents used in mouthwashes such as chlorhexidine, xylitol, silver diamine fluoride and delmopinol lack selectivity, affecting both pathogenic and commensal beneficial species alike giving rise to undesired side effects such as vomiting, diarrhea, addiction, or teeth discoloration⁹. In addition, the biofilm nature of cariogenic bacteria makes it resistant to traditional anti-microbial treatments¹⁰. A few preventive and therapeutic strategies are under investigation by targeting different virulent determinants of S. mutans¹¹. However, small molecules derived from natural products that possess antibacterial activities and antibiofilm properties are randomly identified inhibitors that lack selectivity towards pathogenic biofilms and the in vivo applications of these inhibitors are unclear¹².

Dental plaque consists of more than 700 bacterial species living in complex communities called biofilms⁴. It is initiated by the attachment of commensal streptococci such as Streptococcus sanguinis and Streptococcus gordonii to the saliva-coated tooth surface, which then engage in developing intra-and inter-species bacterial interactions¹³. Under disease conditions, the delicate balance between commensal and pathogenic members of the plaque bacteria is disturbed, leading to an overgrowth of pathogenic species¹⁴. Streptococcus mutans has been implicated as the major etiological agent in the initiation and propagation of this disease¹⁵. The formation of tenacious biofilms is the hallmark of S. mutans induced cariogenesis. Therefore, the studies aimed at developing dental caries treatments should focus on identifying selective inhibitors of biofilms that do not affect the growth of oral commensal bacteria.

Major virulence factors of S. mutans that significantly contribute to its ability to form cariogenic biofilm are its extracellular glucosyl transferases (Gtfs)¹⁶. Most strains of S. mutans harbor three distinct gtf genes expressing different Gtf activities. The genes gtfB and gtfD produce GtfB and GtfD enzymes respectively and synthesize predominantly water-insoluble and soluble glucans¹⁷ correspondingly, while gtfC, encodes for GtfC, an enzyme that synthesizes both water-insoluble and soluble glucans¹⁸. S. mutans GtfB and GtfC are essential for glucan synthesis, bacterial colonization and cariogenesis. Therefore, small molecules inhibitors of S. mutans Gtfs^{19, 20, 21} have potential application in treating and preventing dental caries.

Many anti-biofilm agents display poor efficacy within the oral cavity due to poor solubility, inability to penetrate biofilms and lack of ability to retain in the locally infected areas. Given these challenges, antibacterial nanoparticles have generated recent interest due to their potential applications in anti-caries research. Examples of these are silver nanoparticles in the prevention of dental caries²², farnesol and myricetin co-loaded nanoparticles to inhibit biofilms²³, pH responsive materials to deliver farnesol²⁴, porous silicon microparticles to mitigate cariogenic biofilm²⁵, ferumoxytol nanoparticles²⁶, poly(ethylenimine)²⁷ and chitosan nanoparticles²⁸ with strong antibacterial activity against S. mutans. Several nano systems for controlled release of anti-caries drugs have also been explored including mesoporous silica nanoparticle²⁹, liposome³⁰, halloysite nano-tube³¹, polyamidoamine³² and dextran-coated Iron oxide nanoparticles (nanozymes)³³. Despite the flurry of these recent studies, none of these agents are translated for clinical use as their in vivo efficacies are either modest or not proven.

Under physiological conditions, the human salivary system maintains a healthy pH range of 6.0–7.5 in the oral cavity³⁴ using three buffer systems: 1) bicarbonate, 2) phosphate, and 3) salivary proteins³⁵. A salivary pH below 5.5 is potentially harmful to the hard and soft tissues in the oral cavity³⁶. Under pathogenic oral conditions, biofilms ferment dietary carbohydrates to produce acidic byproducts such as lactic acid, which decreases the pH and causes the demineralization of tooth enamel³⁷. Therefore, a drug that can specifically inhibit the biofilm delivered into the oral cavity in a pH-responsive manner would be highly desirable. Since the pH level in the oral cavity is critical for the demineralization of tooth enamel, our efforts were focused on developing a novel drug delivery system with built-in pH-sensitivity for the delivery of biofilm inhibitors as an anti-caries treatment.

Our recent studies aimed at developing selective small molecule inhibitors of S. mutans biofilm targeted the S. mutans’ surface enzymes, Gtfs^{19, 20, 21}. These studies have resulted in the identification of two potent lead compounds G43 and IIIC5 (Fig. 1)¹⁹. The goals of present study were to improve the solubility of the lead biofilm inhibitor IIIC5, to encapsulate the optimized lead in pH-responsive hydrogel microparticles and to explore its biofilm and growth inhibitory activities in vitro and anti-virulence activities in vivo.

Figure 1:

Chemical structures of G43 and IIIC5.

RESULTS AND DISCUSSION.

Design and synthesis of biofilm inhibiting aurone compounds.

Our initial efforts to prepare pH-responsive hydrogel encapsulated biofilm inhibitors using compounds G43 or IIIC5 did not yield the expected results due to the low solubility (10–25 μg/mL) of these inhibitors. Therefore, efforts were made to modify the structure of IIIC5 to improve solubility. Specifically, analogs of IIIC5 were prepared by substituting the benzofuran ring with a structurally similar aurone ring, removing the nitro, amide, and ester groups, and by introducing multiple hydrophilic OMe or OH groups on the phenyl ring. This study resulted in the identification of several aurone derivatives (MA1–6 and HA2–6, Fig. 2), of which the hydroxyaurones were found to have the desired solubility required for the hydrogel encapsulation while maintaining the potency and selectivity of biofilm inhibition.

Figure 2:

Methoxy and hydroxy aurones

Aurones are a class of organic compounds that are gaining interest in medicinal chemistry due to their biological activities and presence in natural products³⁸. Aurone natural products play an important role in the pigmentation of flowers and fruits³⁹. Their reported bioactivities range from antifungal activity⁴⁰, antifeedant activity⁴¹, tyrosinase inhibition⁴², and antioxidant activity⁴³. In vitro antimicrobial activities of aurones and chalcones are widely reported⁴⁴. Biosynthetically, aurones are derived from chalcones³⁹. Therefore, we took a biomimetic synthetic approach (Scheme 1) to generate a small library of aurones (Fig. 2), which includes one aurone derivative with an unsubstituted phenyl ring (MA1), five methoxy substituted aurones (MA2–6) and five hydroxy substituted aurones (HA2–6). These aurones were prepared from 2-hydroxychalcones (3a-f), which in turn were prepared by the Claisen-Schmidt aldol condensation⁴⁵ of the benzaldehydes (2a-f) and 2-hydroxyacetophenone (1) in the presence of KOH in ethanol in 37–90 % yield. Cyclization of chalcones (3a-f) in the presence of Hg(OAc)₂ in anhydrous pyridine afforded the aurones (MA1–6) in 79–100 % yield. Methyl groups in methoxyaurones (MA2–6) were then removed by treatment with BBr₃ in anhydrous CH₂Cl₂ to afford the hydroxyaurones (HA2–6) in 80–86 % yield.

Scheme 1:

Synthesis of substituted aurones.

One of the goals for this study was to improve the solubility of the lead compound, so solubility of MA1–6 and HA2–6 were determined as reported⁴⁶ (Table 1). As expected, the majority of aurone derivatives had better solubility than the lead compound IIIC5 (25 μg/mL)¹⁹. Among the aurones, hydroxy aurones were found to be more soluble than methoxy aurones and a trend of increasing solubility was observed with the increase in the number of hydroxy groups on the phenyl ring. The hydroxyaurone, HA5 with 2,4,5-trihydroxyphenyl ring was found to be the most soluble analog with the solubility of 120.09 μg/mL. A close analog, HA6 with 3,4,5-trihydroxyphenyl ring had the next highest solubility (90.77 μg/mL). The monohydroxyphenyl analog, HA2 showed the lowest solubility (18.93 μg/mL) among the hydroxyaurones. Methoxyaurones displayed a similar trend of increase in solubility with the increase in number of methoxy groups. The least soluble methoxyaurone was found to be the monomethoxy analog, MA2 with the solubility of 16.23 μg/mL. Trimethoxy aurone analogs, MA5 and MA6 were found to be the most soluble methoxyaurone analogs with the solubilities of 42.36 μg/mL and 44.68 μg/mL, respectively. The only exception to this trend was the 3,5-dimethoxy analog, MA4 which showed lower solubility of 18.97 μg/mL compared to 3,4-dimethoxy analog, MA3 (36.26 μg/mL). The aurone analogs that displayed lower solubility than IIIC5¹⁹ are MA1, MA2, MA4 and HA2.

Table 1:

Reaction yields, solubility and biofilm inhibition profiles of 3a-f, MA1–6, and HA2–6.

Compd No	R2 Group	R1 Group	Yielda (%)	Solubilityb (μg/mL)	Biofilm IC50c (μM)
IIIC5 19		NA	NA	25 ± 0.00	2.70 ± 0.09
3a		H	62	-d	>300e
3b		4-OMe	37	-d	>300e
3c		3,4-di-OMe	53	-d	>300e
3d		3,5-di-OMe	78	-d	>300e
3e		2,4,5-tri-OMe	90	-d	>300e
3f		3,4,5-tri-OMe	85	-d	180.80 ± 0.65
MA1		H	78	8.51 ± 2.00	33.61 ± 0.53
MA2		4-OMe	93	16.23 ± 0.41	107.80 ± 0.65
MA3		3,4-di-OMe	100	36.26 ± 1.56	49.40 ± 4.79
MA4		3,5-di-OMe	89	18.97 ± 2.33	>300e
MA5		2,4,5-tri-OMe	99	42.36 ± 2.86	52.81 ± 7.42
MA6		3,4,5-tri-OMe	89	44.68 ± 0.87	>300e
HA2		4-OH	89	18.93 ± 1.41	18.79 ± 2.36
HA3		3,4-di-OH	92	81.56 ± 2.90	30.67 ± 2.28
HA4		3,5-di-OH	80	76.50 ± 0.57	94.22 ± 2.18
HA5		2,4,5-tri-OH	83	120.09 ± 1.73	6.42 ± 0.61
HA6		3,4,5-tri-OH	86	90.77 ± 0.48	18.92 ± 0.39
G43 21	NA	NA	NA	NA	6.28 ± 0.58
Salicylic acid 47	NA	NA	NA	1880 ± 30	NA

^a)Isolated yield and the compounds are fully characterized with ¹H-NMR, ¹³C-NMR and HRMS;

^b)Solubility in water containing 1 % DMSO determined by UV spectroscopy;

^c)S. mutans UA159 were co-incubated with the compounds at various concentrations and biofilm formation was measured at OD562 using an established crystal violet protocol⁴⁸. IC₅₀ values represent the means ± standard error mean (SEM) from three independent experiments;

^d)Not determined;

^e)Highest concentration tested.

Inhibition of S. mutans UA159 planktonic growth.

In order to identify inhibitors of cariogenic biofilm without affecting the growth of oral bacteria, we first evaluated the effects of compounds 3a-e, MA1–6 and HA2–6 on S. mutans planktonic growth at a single concentration of 50 μM⁴⁸. No significant inhibition of planktonic growth was observed between the control group and treated groups for all chalcone derivatives, 3a-f (Fig. 3A). Methoxyaurones, MA2–6 were found to be slightly more bactericidal than chalcones showing 25–40 % planktonic growth inhibition (Fig. 3B). The aurone analog with unsubstituted phenyl ring (MA1) showed the highest bactericidal activity with 80 % inhibition of the planktonic growth. Some of the hydroxyaurones were more bactericidal than chalcones and methoxyaurones with HA2, HA3 and HA4 showing 60 %, 40 % and 30 % inhibition, respectively. Two hydroxyaurones, HA5 and HA6 did not inhibit the planktonic growth of S. mutans at 50 μM and appeared to be promising lead compounds (Fig. 3C) for further evaluation.

Figure 3:

Planktonic growth inhibitory activities of chalcones (3a-f), methoxyaurones (MA1–6), and hydroxyaurones (HA2–6). A) S. mutans UA159 were co-incubated with 50 μM of chalcones 3a-f and the planktonic growth was measured at OD₄₇₀. B) S. mutans UA159 were co-incubated with 50 μM of methoxy aurones, MA1–6 and the planktonic growth was measured at OD₄₇₀. C) S. mutans UA159 were co-incubated with 50 μM of hydroxyaurones, HA2–6 and the planktonic growth was measured at OD₄₇₀. Each experiment was repeated three times with triplicate microwells for each compound. Statistical significance was tested with one-way ANOVA. p<0.0001.

Inhibition of S. mutans UA159 biofilms.

Initial screening of compounds 3a-f, MA1–6 and HA2–6 in a single species S. mutans biofilm assay was carried out at a single treatment dose of 50 μM. Members of all three series of compounds were effective in inhibiting biofilms with hydroxyaurones exhibiting most pronounced activity compared to methoxyaurones and chalcones (Fig. 4). More importantly, all compounds showed varying degrees of selectivity towards inhibition of biofilm as opposed to growth. Chalcones, 3a-f were generally less active compared to aurones (Fig. 4A). The most active chalcone derivative, 3f exhibited 40 % biofilm inhibition and no growth inhibition at 50 μM. The most potent methoxyaurone, MA5, exhibited 60 % biofilm inhibition (Fig. 4B). However, this compound also inhibited 30 % of bacterial growth at 50 μM making it a less selective biofilm inhibitor. The other methoxyaurones, MA1, MA2, MA4 and MA6 were relatively less active displaying only 20–40 % biofilm inhibition, while MA3 was inactive at this dose. Overall, hydroxy aurones were better biofilm inhibitors than chalcones and methoxyaurones with derivatives, HA2, HA5 and HA6 showing more than 95 % inhibition and HA3 showing about 80 % inhibition of biofilms (Fig. 4C). Among the most active hydroxyaurones, 4-hydroxy analog, HA2 inhibited bacterial growth by 70 % at the treated dose, making it a less selective biofilm inhibitor (Fig. 3C). The 3,5-dihydroxy aurone analog, HA4 did not show significant biofilm inhibition. The 2,4,5-trihydroxy and 3,4,5-trihydroxy analogs, HA5 and HA6, respectively were found to be the most active hydroxyaurone analogs with more than 95 % biofilm inhibition and no effect on growth at 50 μM, making them the most active and selective biofilm inhibitors from this screening (Fig. 4C).

Figure 4:

Biofilm inhibitory activities of chalcones (3a-f), methoxyaurones (MA1–6) and hydroxyaurones (HA2–6). A) S. mutans UA159 were co-incubated with 50 μM of chalcones 3a-f and biofilm formation was measured at OD₅₆₂ using the crystal violet protocol. B) S. mutans UA159 were co-incubated with 50 μM of methoxyaurones, MA1–6 and biofilm formation was measured at OD₅₆₂ using the crystal violet protocol. C) S. mutans UA159 were co-incubated with 50 μM of hydroxyaurones HA2–6 and biofilm formation was measured at OD₅₆₂ using the crystal violet protocol. Each experiment was repeated three times with triplicate microwells for each compound. Statistical significance was tested with one-way ANOVA. p<0.0001.

Inhibition of commensal Streptococci biofilms by hydroxyaurones.

To determine the selectivity of the hydroxyaurones (HA2–6) toward S. mutans biofilm formation over the biofilms of commensal species, we evaluated effects of these compounds on the biofilm formation by two oral commensal Streptococci bacteria: S. gordonii and S. sanguinis. At 50 μM concentration, S. gordonii biofilm formation was inhibited by 40–60 % (Fig. 5A), while S. sanguinis biofilm was inhibited by 30–40 % (Fig. 5B). However, these effects were less pronounced than their effects on S. mutans biofilm. For example, compounds HA5 and HA6 displayed about 95 % inhibition of S. mutans biofilm at 50 μM (Fig. 4C). These effects were also comparable to the control Gtf inhibitor G43 reported from our lab previously. In addition, a side-by-side comparison of biofilm inhibitory effects of 25 μM HA5 inhibited 80 % of S. mutans biofilm while it did not significantly reduce S. Sanguinis biofilm and inhibited about 20 % of S. gordonii biofilm (Fig. 8G). Overall, this data suggests that hydroxyaurones have a high degree of selectivity towards inhibiting pathogenic biofilms compared to commensal biofilms.

Figure 5:

Inhibitory activities of hydroxyaurones (HA2–6) against commensal biofilms. A) S. gordonii DL1 were co-incubated with 50 μM of hydroxyaurones, HA2–6 or G43 and biofilm formation was measured at OD₅₆₂ using the crystal violet protocol. B) S. sanguinis SK36 were co-incubated with 50 μM of hydroxyaurones, HA2–6 or G43 and biofilm formation was measured at OD₅₆₂ using the crystal violet protocol. Each experiment was repeated three times with triplicate microwells for each compound. Statistical significance was tested with one-way ANOVA. p<0.0001.

Figure 8.

A) Optical images of empty (PMAA)₅ hydrogels microparticles. B) HA5-loaded hydrogel HEBI and HA5 in methanol (insert B). C) Atomic Force Microscopy (AFM) topography images of a tooth surface with height of 280 nm. D) AFM image after (PMAA)₅ hydrogel adsorption, cubical hydrogel particles are clearly seen sticking to the tooth surface. E) Amplitude error image of empty (PMAA)₅ hydrogels dried on the surface of a tooth. Scan size is 20 μm2 in both images, the height (z)-scale is 1.7 μm. F) S. mutans UA159 and two bacterial commensal species S. gordonii DL1 or S. sanguinis SK36 were co-incubated with HA5 or HEBI at 25 μM and their growth was measured at OD₄₇₀. G). S. mutans UA159, S. gordonii DL1 or S. sanguinis SK36 were co-incubated with 25 μM of HA5 or HEBI and biofilm formation was measured at OD₅₆₂ using the crystal violet protocol. Each of the biofilm and growth assays were conducted in triplicate and statistical significance was tested with one-way ANOVA. p<0.0001.

Considering the potential of methoxyaurones and hydroxyaurones for further development, their biofilm inhibitory activities were further characterized in serial dilutions and IC₅₀ values were determined. The hydroxyaurones were found to have lower IC₅₀ values compared to the corresponding methoxyaurones (Table 1). Among methoxyaurones, 3,4-dimethoxyaurone, MA3 was found to be the most active analog with an IC₅₀ value of 49.40 μM. The 2,4,5-trimethoyaurone, MA5 had a similar IC₅₀ value of 52.81 μM and the 4-methoxyaurone, MA2 had an IC₅₀ value of 107.80 while 3,4-dimethoxy and 3,4,5-trimethoxy aurones were inactive. Interestingly, the unsubstituted aurone, MA1 was more potent than all methoxy aurones with an IC₅₀ value of 33.61 μM. However, MA1 also displayed about 80 % inhibition of S. mutans growth at 50 μM, suggesting that its observed biofilm inhibition may be arising from its bactericidal activity.

Two of the hydroxyaurones, 2,4,5-trihydroxy aurone (HA5) and 3,4,5-trihydroxy aurone (HA6) were found to be the most active derivatives with IC₅₀ values of 6.42 μM and 18.92 μM respectively. The 3,4-dihydroxyaurone, HA3 and 3,5-dihydroxyaurone, HA4 were found to be less active with IC₅₀ values of 30.67 μM and 94.22 μM, respectively. Among these, HA4 with no OH at the 4-position, was less active than HA3 with an OH group at 4-position. Interestingly, the monohydroxy analog, HA2 with an OH group at 4-postion was found to be more active than the dihydroxyaurones, HA3 and HA4. It should be noted that both of our most active analogs HA5 and HA6 also contained an OH group at the 4-position, indicating the importance of the 4-OH group for the biofilm inhibitory activities of hydroxyaurones. This observation is further supported by our co-crystal structure of HA5 in the GtfB active site, showing that the two oxygen atoms at the 4,5-position of the 2,4,5-trihydroxyphenyl moiety interacted with the key amino acid residues in the active site through the coordination with a conserved Ca²⁺ ion (Fig. 7). Of all the aurone analogs synthesized, 2,4,5-trihydroxyaurone, HA5 (Fig. 6E) was selected as our lead compound for further analysis and encapsulation studies based on its potent biofilm inhibition, lack of growth inhibition and improved solubility.

Figure 7:

A) High resolution X-ray co-crystal structure (PDB ID: 8FG8) of the inhibitor HA5 with the catalytic domain of GtfB. Inhibitor HA5 and BTB are displayed as green sticks. B) Expansion of the GtfB binding site showing the binding mode of HA5. C) Key active site interactions of HA5 with active site residues (grey sticks) along with its H-bong interactions with water (Wat) molecules and calcium (Ca²⁺) ion depicted as yellow dotted lines. Inhibitor HA5 is depicted as light blue sticks and BTB is depicted as lavender sticks.

Figure 6:

Biofilm inhibitory activities of compound HA5. A) S. mutans UA159 were co-incubated with HA5 at various concentrations and biofilm formation was measured at OD₅₆₂ using the crystal violet protocol. B) Gtfs precipitated from S. mutans culture were co-incubated with HA5 at various concentrations and the glucan production was quantified using cascade blue staining and subsequent image processing with ImageJ. C) Representative fluorescence microscopy images of UA159 biofilms after 16 h of treatment with various concentrations of HA5. Bacterial cells were stained with Syto-9 (green, panel I); glucans were stained with Cascade Blue–dextran conjugated dye (blue, panel II); eDNA was stained with propidium iodide (red, panel III) and a merged image of all three staining images (panel IV). D) S. mutans UA159, S. gordonii DL1 and S. sanguinis SK36 were co-incubated with HA5 at 50 μM and 100 μM and their growth were measured at OD₄₇₀. E) Chemical structure of HA5. Each of the biofilm, glucan and growth assays were conducted in triplicate and statistical significance was tested with one-way ANOVA. p<0.0001.

HA5 inhibits S. mutans UA159 biofilms, glucan production and eDNA levels.

The antibiofilm activities of HA5 were further investigated by fluorescence microscopy imaging. Compound HA5 displayed a dose-dependent inhibition of S. mutans biofilm as shown in Fig. 6A. Staining of bacterial cells within biofilms with Syto-9 showed significant reduction in biofilms at 5 μM of HA5 and a complete inhibition at 50 μM of HA5 (Fig. 6C, Panel-I). The presence of glucans, which were stained with Cascade Blue-dextran conjugated dye, was significantly reduced at 5 μM of HA5 and no glucan formation was evident at 50 μM of HA5 (Fig. 6C, Panel II). In addition, propidium iodide was used to determine the presence of extracellular DNA (eDNA) in S. mutans biofilms. Again, there was a noticeable reduction of eDNA at 5 μM of HA5 and almost complete absence of eDNA at 50 μM of HA5 (Fig. 6C, Panel III). These findings reaffirm that HA5 inhibited S. mutans biofilms by preventing the synthesis of glucans and minimizing the presence of eDNA, two integral biofilm matrix elements crucial for S. mutans biofilm formation.

HA5 inhibits the glucan production of S. mutans UA159 in a dose dependent manner.

The interspecies co-adherence between S. mutans and other microorganisms in the oral cavity is critical for biofilm formation and cariogenicity. Though the mechanisms of such adhesions and co-aggregations are not fully elucidated, it is believed that the extracellular polysaccharide (EPS) matrix of S. mutans has an important role in this process^{49, 50}. It is reported that glucans synthesized by Gtfs when incorporated into the tooth pellicle to provide enhanced binding sites for other microorganisms to form stable and persistent microcolonies, which provides mechanical stability to the EPS matrix^{50, 51}. Therefore, Gtf inhibition assays were performed to assess the ability of HA5 to inhibit the Gtfs and glucan production using a reported procedure and IC₅₀ value was calculated⁵². Compound HA5 exhibited dose dependent inhibition of glucan production by Gtfs with an IC₅₀ value of 10.56 μM (Fig. 6B). These findings reinforce the biofilm inhibitory activity of HA5 and suggest that the compound inhibits biofilm formation by inhibiting glucans production by S. mutans Gtfs.

HA5 does not affect the growth of commensal streptococcal species.

To determine if compound HA5 only selectively inhibits S. mutans biofilms over the growth of S. mutans and oral commensal species, the effects of HA5 on the growth of two representative commensal oral streptococci, S. gordonii and S. sanguinis, along with S. mutans at 50 μM and 100 μM doses were evaluated. As shown in Fig. 6D, compound HA5 did not inhibit the growth of two commensals compared to the control group at these doses that are much higher than its biofilm IC₅₀ value of 6.42 μM. Similarly, the compound did not inhibit S. mutans growth at these doses, suggesting that HA5 selectively inhibited S. mutans biofilms without affecting its growth as well as the growth of commensal species, S. gordonii and S. sanguinis (Fig. 6D).

Structural studies on HA5 in the catalytic domain of GtfB.

Apo crystals of GtfB were obtained in the Index (Hampton Research) A3 crystallization condition using the hanging drop vapor diffusion technique produced reliably large enough (0.1–0.2 mm) tetragonal crystals for GtfB. Apo crystals of GtfB were soaked with HA5 for 5–10 mins (final concentration 1–2 mM; 20 mM stock solution of compound HA5 in H₂O). The GtfB structure complexed with HA5 diffracted to a resolution of 2.35 Å. Diffraction data for the inhibitor structure were collected at 100 K using the hybrid pixel DECTRIS Eiger 16m detector at the SERCAT 22-ID beamline in the Advanced Photon Source (APS), Chicago. For cryoprotection of GtfB crystals 20 % ethylene glycol was added to the crystallization buffer. The collected data were processed using XDS⁵³ for initial indexing, merging, and scaling, and were followed by optimization using Aimless⁵⁴ in CCP4⁵⁵. The data collection statistics are shown in Table 2. The structure for GtfB was solved by molecular replacement using Phaser with a model for GtfB generated by the SWISS-MODEL web server based on the catalytic domain of GtfC (3AIE). Refinement was performed using a combination of Refmac5⁵⁶ in CCP4⁵⁵ and Phenix⁵⁷. MOGUL (CSD release) restraints for compound HA5 were based on the CCDC small molecule database and obtained from the Grade web server. Compound HA5 is unknown to PDB, and therefore, was sketched in ChemAxon and the resulting SMILES notation provided the initial input. Coot was used for all model building⁵⁸ and figures were created with PyMOL (Version 2.5.0, Schrödinger, LLC). Validation of the model quality was performed using Phenix and the wwPDB validation service. The ligand was validated using the same set of MOGUL restraints described above. The conserved Ca²⁺ site in the structure of GtfB was verified with the CheckMyMetal web server. The protein structure has been deposited in PDB. Final refinement statistics are presented in Table 2.

Table 2:

Data collection and refinement statistics for GtfB in complex with HA5

Data collection
Space Group	P4322
Unit cell parameters [Å]	a = b = 150.46, c = 304.95
Resolution [Å]	87.25 – 2.35 (2.39 – 2.35)
Unique reflections	142674 (7062)
Completeness [%]	98.5 (99.2)
Multiplicity	7.4 (7.2)
Rmerge [%]	19.1 (210.7)
Rpim [%]	6.5 (71.2)
CC1/2 (2.39 – 2.35Å)	0.350
CC* (2.39 – 2.35Å)	0.720
I/σ(I)	8.1 (1.1)
Refinement
Resolution [Å]	73.04 – 2.35 (2.41 – 2.35)
No. of reflections	139861 (10356)
Completeness [%]	96.1 (97.6)
Rwork [%]	20.5 (33.6)
Rfree [%]	23.3 (35.5)
Wilson B [Å2]	48.0
Average B-factors [Å2]
Overall	57.9 (13785 atoms)
Protein, Ca2+, HA5, BTB	57.8 (13091 atoms), 55.9 (3 ions), 83.7 (40 atoms), 73.9 (28 atoms)
SO42−, Waters	97.7 (125 atoms), 47.2 (503 atoms)
Rmsd bonds [Å]	0.013
Rmsd angles [°]	1.74
CC (Fo-Fc)	0.95
Ramachandran [%]	95.9 favored (outliers 0.7)
Clash score	3.76
Molprobity score	1.44

Interactions of HA5 within GtfB active site.

To define the underlying mechanism of HA5’s ability to inhibit Gtfs and biofilm, a high-resolution co-crystal structure of HA5 with the catalytic domain of GtfB was resolved. The analysis of GtfB/HA5 co-crystal structure revealed that the inhibitor HA5 was crystallized along with Bis-Tris (BTB), a chemical component of the buffer used in the crystallization studies (Fig. 7A–B). Inhibitor HA5 was found to adopt a classic π-π stacking interaction with Trp491 residue in the GtfB active site (Fig. 7C). In the structure only water-mediated hydrogen bonds are observed for HA5, which differs from the reported in silico docking results with the inhibitor G43, which highlighted hydrogen bond interactions of G43 with the three active site residues Asp451, Glu489 and Asp562¹⁹. In the crystal structure of HA5, the inhibitor makes only hydrophobic contact with these residues (Fig. 7C). For the sucrose to undergo the invertase activity, the glucose molecule will nest within the −1 subsite which consists of residues Arg449, Asp451, Glu489, His561, Asp562 and Tyr890. And the fructose interacts with the +1 subsite that consists of Tyr404, Leu407 and Trp491. These nomenclatures are adopted for sugar binding sites for glycosylhydrolases⁵⁹. The inhibitor HA5 in this crystal structure binds to subsites +1 and +2 that include the residues Asn511, Arg514, and Asp567. Inhibitor HA5 makes hydrophobic interactions with a total of 15 interface residues in these two subsites. 50 % of its solvent accessible surface area is buried in the protein-ligand interactions. In chain B of GtfB, two oxygen atoms at the 4,5-position of the 2,4,5-trihydroxyphenyl moiety of the inhibitor HA5 coordinated with a conserved Ca²⁺ ion and extends its interaction with Asp451 and Glu489 (Fig. 7C). The 4-OH group on the benzene ring of HA5 interacts with Tyr404. In chain A of GtfB, the 2,4,5-trihydroxyphenyl moiety is rotated around the methylidene atom with respect to the benzofuran ring system allowing a hydrogen bond with BTB buffer molecule (buffer of the crystallization condition) within the active site. The result of finding BTB in the GtfB active site is not surprising since Tris as an ethanolamine derivative has been previously reported as a competitive inhibitor of GtfB⁶⁰. Binding of the BTB in this structure occurs in −1 subsite and overlaps with proposed binding of the glucosyl moiety of sucrose. The BTB molecule provides 11 hydrogen bonds to protein residues within the active site. Placement of HA5 and BTB in both subunits were not only verified in 2mFo-DFc maps at 1 sigma contour level but also in Polder difference omit maps at 5 sigma contour levels (Supplementary Fig. S1).

We realize that HA5 is a polyhydroxy compound that contains a Michael acceptor functionality which raises concerns about non-specific and covalent binding. However, it is unlikely that HA5 is influenced by these mechanisms because our HA5/GtfB co-crystal structure clearly shows its binding in the catalytic site of GtfB with specific interactions with the Ca²⁺ ion and with active site residues and it does not show any covalent bond to its Michael acceptor site. To further validate that HA5 is not a Michael acceptor, the Gtf inhibition IC₅₀ values for HA5 in the presence and absence of a nucleophilic reagent, beta-mercaptoethanol (BME, 1 mM) have been determined and shown that BME doesn’t reduce the Gtf inhibitory activity (7.84 μM vs 10.56 μM)⁶¹. In addition, the Gtf inhibition IC₅₀ values for HA5 in the presence and absence of a detergent Triton-X-100 have been determined to show that it is not a non-specific inhibitor⁶². Triton-X-100 did not reduce the Gtf inhibitory activity (5.16 μM vs 10.56 μM) of HA5 suggesting that the observed Gtf inhibition is not due to non-specific binding.

Hydrogel encapsulated biofilm inhibitors.

Hydrophilicity, the ease of chemical modification and structural stability of hydrogel matrices ensure excellent biocompatibility and versatility for its use in biomedical applications. Poly(methacrylic acid) [PMAA] hydrogel is an excellent platform for the pH-triggered drug delivery of the biofilm inhibitors as these respond to varying pH due to the existence of ionizable pendant groups (e.g. −COOH and −NH₂) in the network. In our previous studies, PMAA hydrogels have been prepared by layer-by-layer (LbL) assembly of hydrogen-bonded polymers of PMAA and poly(N-vinylpyrrolidone) (PVPON). The PMAA and PVPON layers were alternatingly adsorbed onto surfaces of porous inorganic microparticles of manganese oxide, followed by chemical crosslinking of PMAA with ethylenediamine and dissolution of the manganese oxide template microparticles^63–65. The nanoscale multilayers of chemically crosslinked PMAA result in the interconnected porous hydrogel structure, which provides excellent drug loading capacity. Besides, the pH-responsiveness of the hydrogel can be easily tuned during particle formation by using pH-sensitive cross-linkers⁶⁶. We have recently demonstrated the biocompatibility and degradability of hydrogel biomaterial in the delivery of small-molecule drugs⁶⁵.

Encapsulation of HA5 inside (PMAA)₅ hydrogels microparticles.

Compound HA5 was encapsulated in the (PMAA)₅ hydrogel cubes through post-loading by soaking the hydrogels in 5 mg/mL solution of HA5 in methanol for 48 h in the dark (Figs. 8A–B). The free, non-encapsulated HA5 was removed from particle solution by rinsing with HEPES buffer (pH = 7.4) five times using centrifugation at 5000 rpm for 10 min. The HA5 quantification was carried out with UV-visible spectroscopy (NanoDrop One C, ThermoFisher) at λ = 448 nm using an HA5 calibration curve. The drug solution was analyzed before and after the exposure to the hydrogel particles and the differences in the absorbance spectra were used to determine the loading of the drug into the hydrogel network. The loading capacity was found to be 5.5 × 10⁻³ ng of HA5 per particle. To demonstrate the tooth adhesion of (PMAA)₅ hydrogel microparticles, a drop of the hydrogel particle dispersion was placed on the tooth surface and dried at room temperature for 10 min in a Petri dish and morphology of the hydrogels were analyzed using atomic force microscopy (AFM NTEGRA II microscope: NT-MDT) imaging. Freshly extracted, intact third molars with flat surfaces obtained from Dr. Nathaniel Lawson’s lab (UAB School of Dentistry, IRB-300001291) were used in these studies. The AFM silicon probes NSG30 (NT-MDT, resonance frequency 240–440 kHz, force constant 22–100 N m^–1, tip radii is 10 nm, scan rate is 0.5 Hz) were used for imaging the tooth surfaces in tapping mode before and after hydrogel adhesion. The AFM image shows that the bare tooth surface displays natural topography (Fig. 8C) with height of 280 nm. After hydrogel addition, the cubical hydrogel particles are seen to adhere to the tooth surface (Figs. 8D–E). The height of the dried hydrogel cubes was determined using section profiles, which indicated an average particle height of 1.3 ± 0.2 μm. The hydrogel cubes decreased in size compared to their size in solution due to the hydrogel shrinkage upon drying⁶⁵.

Inhibition of biofilms and planktonic growth of by HEBI.

Effects of HEBI and HA5 on biofilms and planktonic growth of S. mutans and two commensal Streptococci, S. gordonii and S. Sanguinis were evaluated at a single treatment dose of 25 μM. HEBI inhibited about 85–90 % of S. mutans biofilms, which is comparable to 90 % inhibition by HA5 (Fig. 8G). HEBI did not significantly inhibit the biofilms of commensal species S. gordonii and S. sanguinis at this concentration. As expected, based on our previous data, HEBI or HA5 did not affect the planktonic growth of S. mutans, or the commensal species S. gordonii and S. sanguinis at the treatment concentration of 25 μM (Fig. 8F).

Reduction of S. mutans virulence in vivo by HA5 or HEBI.

The effects of compound HA5 and HEBI on S. mutans virulence were evaluated using a well-established gnotobiotic rat model of dental caries⁶⁷. Hydrogel microparticles with no drug were used to ensure that the observed anti-virulence activity observed with HEBI was not related to the hydrogel material. The standard NaF (250 ppm) was included as a positive control. A (vehicle + infection only) group was included as a negative control. All rats in the experimental groups and control groups were colonized with S. mutans UA159. A 4-week treatment of S. mutans UA159 infected gnotobiotic rats with 100 μM of HA5 or HEBI resulted in significant reduction in buccal and sulcal caries scores compared to control groups. Similar reductions in caries scores were also observed in proximal enamel caries scores (Table 3). We were unable to evaluate the effect of the treatment on proximal dentinal scores as there were no significant proximal dentinal lesions for the control and treated groups in this study. In comparison, the group treated with hydrogel (no drug) did not show any inhibition compared to the control group suggesting that the hydrogel as such has no antivirulence activity (Table 3). The observed reduction in caries scores by HA5 and HEBI were similar with HEBI displaying slightly better in vivo activity, possibly due to the pH-dependent slow release. The observed reduction in caries scores by HA5 or HEBI is lower than the 250 ppm NaF treatment. However, it should be noted that the concentration of NaF (250 ppm = 5.95 mM) is about 59-fold higher than HA5 (100 μM). At the end of the study, the animals were euthanized, their mandibles excised for microbiological analysis of plaque samples on MS agar plates and BAP and for scoring of caries by the method of Keyes⁶⁸ to determine the bacterial colonization. The effect on bacterial colonization was not significant in HA5 or HEBI treated animals when compared to control group, while the bacterial colonization appears to be slightly reduced in unloaded hydrogel treated rats. This data suggests that HA5 and HEBI are less toxic to bacteria (Table 4). Moreover, the rats treated with the compound HA5 or HEBI did not experience any weight loss over the course of the study in comparison with the control group, suggesting that they are non-toxic (Table 4). Overall, our data suggest that HEBI can release HA5 in the rat’s oral cavity under the acidic conditions of dental caries and the reduction in caries scores produced by HEBI is comparable to what is observed for HA5 treatment alone. These results also indicate that the compound HA5 or HEBI selectively target S. mutans virulence factors; Gtfs and Gtf-mediated biofilm formation, rather than a simple inhibition of bacterial growth and are very effective in inhibiting dental caries in vivo. All in vivo experimental protocols were approved by the University of Alabama at Birmingham Institutional Animal Care and Use Committee (protocol No: IACUC-20047). The methods were carried out in accordance with the relevant guidelines and regulations.

Table 3:

Effect of HEBI or HA5 treatment on S. mutans UA159 induced dental caries.

Treatment Group	Buccal Mean Caries Scores (± SEM)				Sulcal Mean Caries Scores (± SEM)				Proximal Mean Caries Scores (± SEM)
	E	Ds	Dm	Dx	E	Ds	Dm	Dx	E
UA159 untreated	14.6 ± 0.2	10.2 ± 0.4	6.0 ± 0.0	3.2 ± 0.5	25.6 ± 0.2	17.8 ± 0.2	10.4 ± 0.2	4.2 ± 0.2	6.0 ± 0.0
Hydrogel (no drug)	15.0 ± 0.3	9.6 ± 0.8	6.4 ± 0.7	3.8 ± 1.1	27.0 ± 0.5	19.6 ± 0.4	12.4 ± 0.9	6.0 ± 0.5	6.8 ± 0.5
HEBI (100 μM)	7.4 ± 1.1	5.2 ± 1.2	2.4 ± 0.7	1.4 ± 0.6	16.0 ± 0.6	12.6 ± 0.2	6.8 ± 0.7	4.0 ± 0.0	3.6 ± 0.4
HA5 (100 μM)	11.0 ± 0.3	6.0 ± 0.4	3.8 ± 0.2	2.0 ± 0.7	22.0 ± 0.6	16.2 ± 0.4	8.8 ± 0.9	4.4 ± 0.9	4.0 ± 0.0
NaF (250 ppm)	5.0 ± 0.7	2.4 ± 0.8	1.0 ± 0.8	0.0 ± 0.0	14.2 ± 0.7	11.2 ± 0.4	4.6 ± 0.2	1.0 ± 0.3	1.2 ± 0.8

Enamel (E); Dentinal slight (Ds); Dentinal moderate (Dm); Dentinal extensive (Dx); Proximal dentinal scores are not included as there were no significant proximal dentinal lesions for the control and treated groups in this study.

Table 4:

Effect of HEBI or HA5 treatment on S. mutans UA159 CFU and the body weight of the animals.

Treatment Group	CFU/mL (x 106)		Animals
	MS	BAP	Weight (g)	Number
UA159 untreated	4.2 ± 1.4	5.6 ± 1.6	141 ± 13	5
Hydrogel (no drug)	1.6 ± 0.5	2.4 ± 0.7	130 ± 9	5
HEBI (100 μM)	5.2 ± 1.4	5.5 ± 1.4	145 ± 12	5
HA5 (100 μM)	2.5 ± 0.8	3.0 ± 1.0	137 ± 13	5
NaF (250 ppm)	3.4 ± 0.5	4.0 ± 0.9	161 ± 13	5

Colony Forming Unit (CFU); Mitis Salivarius (MS); Blood Agar Plates (BAP).

CONCLUSIONS.

In conclusion, we have developed novel small-molecule inhibitors of S. mutans glucosyl transferases as selective biofilm inhibitors that do not affect the growth of oral commensal bacteria. The solubility and biofilm inhibitory activities of the lead compound were optimized for drug-encapsulation. The optimized lead compound, HA5 inhibited S. mutans biofilm with an IC₅₀ value of 6.42 μM without affecting its growth. Compound HA5 was further evaluated for its effect on the growth of oral commensal bacterial species S. gordonii and S. sanguinis and showed that it does not inhibit the growth of S. gordonii and S. sanguinis at 100 μM, which is 14-fold higher dose than its biofilm IC₅₀ value. The binding of HA5 to the glucosyl transferase, GtfB has been demonstrated by resolving a high-resolution X-ray co-crystal structure of HA5 with the catalytic domain of GtfB and mapped out its active site interactions. Compound HA5 inhibited S. mutans Gtfs and glucan production with an IC₅₀ value of 10.56 μM in a Gtf inhibition assay. Compound HA5 was encapsulated into pH-responsive hydrogel microparticles to generate a hydrogel encapsulated biofilm inhibitor (HEBI), which displayed selective inhibition of S. mutans biofilm similar to HA5. The effects of HA5 or HEBI on the biofilm on commensal species, S. gordonii and S. sanguinis were minimal at 25 μM. A 4-week treatment of S. mutans UA159 infected gnotobiotic rats with 100 μM of HA5 or HEBI resulted in significant reduction in buccal, sulcal, and proximal dental caries scores compared to control groups demonstrating their antivirulence activities in vivo without affecting the bacterial colonization significantly. The rats treated with the HA5 or HEBI did not experience any weight loss over the course of the study in comparison with the control group, suggesting that the compound and material are non-toxic. Overall, our in vivo data suggests that HEBI can release HA5 in the rat oral cavity under the acidic conditions of dental caries infection and the reduce dental caries and the results are comparable to what is observed for HA5 treatment alone. Overall, the results of this study suggest that compound HA5 or HEBI selectively targeted S. mutans Gtfs and Gtf-mediated biofilm formation, rather than a simple inhibition of bacterial growth, demonstrating the potential of this compound and material to be developed further as novel dental caries treatments.

EXPERIMENTAL.

General considerations.

¹H-NMR and ¹³C-NMR spectra were recorded on Bruker Avance Neo 400 and Avance II 700 spectrometers using TMS or appropriate solvent signals as internal standard. The chemical shift values are given in parts per million (ppm) relative to the internal standard used and the coupling constants (J) are given in hertz (Hz). High resolution mass spectra (HRMS) were recorded using Waters AutoSpec-Ultima^™ NT magnetic sector mass spectrometer with Electron Impact (EI) Ionization source. The mass analyzer is an electric-magnetic-electric (EBE) sector (a double focusing sector). Anhydrous solvents used for reactions were purchased in Sure-Seal^™ bottles from Aldrich chemical company. Other chemical reagents were purchased from Aldrich or Fisher chemical companies and used as received. Reactions were monitored with thin layer chromatography (TLC), which was done on silica gel plates with fluorescent indicator (Silicycle, silica gel, UV254, 25 μm plates). The TLC spots were observed under UV light with the wavelengths 254 nm and 365 nm. The reaction mixtures were purified by column chromatography using Si gel (32–63μm) from Dynamic Absorbent, Inc. Melting points were determined on a Mel-Temp II melting point apparatus and are uncorrected. All tested compounds have ≥95 % purity as determined by HPLC. HPLC traces were obtained using Shimadzu SPD-M20A. Solubility of compounds were determined by UV-spectroscopy method using Agilent Cary 60 UV-Vis spectrophotometer. HPLC analysis of the final compounds were conducted using Kinetex 5 μm C18 100 Å, LC Column 150 × 4.6 mm, compound = 3 mM, 20 μL injection, solvent: mobile phase buffer, Conditions: 60 % MeCN / 40 % H₂O / 0.1 % Formic acid (isocratic), HPLC method 0–10 min, Signals were analyzed using a 254 nm UV detector. A chromatogram of Mobile Phase Buffer (20 μL) was obtained for comparison.

Poly(ethyleneimine) (PEI, average Mw 25000), ethylenediamine (EDA), manganese sulfate monohydrate, ammonium bicarbonate and 1-Ethyl-3-(3-(dimethylamino)propyl)-carbodiimide hydrochloride were purchased from Sigma-Aldrich. Poly(methacrylic acid) (PMAA, average M_w 22000 g mol⁻¹, Ð = 1.3) were purchased from Fisher Scientific. Ultrapure de-ionized (DI) water with a resistivity of 18.2 MΩ-cm at 25 °C was used in all experiments. Monobasic and dibasic sodium phosphate (Fisher Scientific) were used for preparation of polymer and buffer solutions. Poly(N-vinylpyrrolidone) (PVPON, M_w 10000 g mol⁻¹) was from Sigma-Aldrich. Slices of human teeth were provided by Dr. Nathaniel Lawson (UAB School of Dentistry, IRB-300001291) and used as received.

The bacterial strains, S. mutans UA159, S. gordonii DL1, and S. sanguinis SK36 were inoculated statically at 37 °C under 5 % CO₂ in Todd Hewitt Broth (THB) for 24 h. The cultures were then diluted with fresh THB (1:5) and reinoculated until optical density at 470 nm (OD₄₇₀) reached 1. The optical density was read using BioTek 800TS microplate reader at 470 nm for bacterial growth and 562 nm for biofilm stained with crystal violet. Data was plotted in Graphpad Prism9.

General procedure for the synthesis of chalcones (3a-f).

To a solution of the 2-hydroxyacetophenone 1 (1 mmol) and benzaldehyde 2a (1 mmol) in EtOH (10 mL), an aqueous solution of KOH (40 %, 1 mL) was added, and the reaction mixture was stirred at room temperature for 12 h. TLC examination (30 % EtOAc in hexanes) indicated the completion of the reaction. The reaction mixture was then poured over crushed ice and acidified the pH to 2 using 1.0 N HCl. The precipitate formed was filtered, washed with copious amounts of water, and dried to obtain the crude product, which was purified on column chromatography over Si gel using 10 % EtOAc in hexanes as eluent to afford clean chalcones 3a-f. All chalcone products were characterized by ¹H NMR, ¹³C NMR and HRMS as follows.

1-(2-Hydroxyphenyl)-3-phenyl-2-propen-1-one (3a):

61.9 % yield, yellow solid; mp. 89–90 °C; ¹H NMR (700 MHz, CDCl₃) δ: 12.84 (s, 1H), 7.94–7.92 (m, 2H), 7.68–7.66 (m, 3H), 7.52–7.49 (m, 1H), 7.45 (t, 3H, J = 3.1 Hz), 7.04 (d, 1H, J = 8.5 Hz), 6.95 (t, 1H, J = 7.6 Hz); ¹³C NMR (700 MHz, CDCl₃) δ: 193.7, 163.6, 145.5, 136.4, 134.6, 130.9, 129.7, 129.0, 128.7, 120.1, 120.0, 118.9, 118.6; HRMS [M-H]-calculated for C₁₅H₁₂O₂ 223.0759, found 223.0763.

1-(2-Hydroxyphenyl)-3-(4’-methoxyphenyl)-2-propen-1-one (3b):

36.5 % yield, yellow solid; mp. 93–95 °C; ¹H NMR (700 MHz, CDCl₃) δ: 12.96 (s, 1H), 7.93–7.89 (m, 2H), 7.63 (d, 2H, J = 8.6 Hz), 7.54 (d, 1H, J = 15.4 Hz), 7.49 (t, 1H, J = 7.7 Hz), 7.02 (d, 1H, J = 8.3 Hz), 6.96–6.93 (m, 3H), 3.86 (s, 3H); ¹³C NMR (700 MHz, CDCl₃) δ: 193.7, 163.6, 162.0, 145.4, 136.2, 130.6, 129.5, 127.3, 120.1, 118.8, 118.6, 117.6, 114.5, 55.5; HRMS [M+H]⁺ calculated for C₁₆H₁₄O₃ 255.1021, found 255.1014.

3-(3’,4’-Dimethoxyphenyl)-1-(2-hydroxyphenyl)prop-2-en-1-one (3c):

53.0 % yield, yellow solid; mp. 115–117 °C; ¹H NMR (400 MHz, CDCl₃) δ: 12.94 (s, 1H), 7.94 (dd, 1H, J = 8.1, 1.6 Hz), 7.89 (d, 1H, J = 15.4 Hz), 7.53 (d, 1H, J = 15.4 Hz), 7.52–7.47 (m, 1H), 7.29–7.26 (m, 1H), 7.18 (d, 1H), 7.03 (dd, 1H, J = 8.3 Hz), 6.97–6.91 (m, 2H), 3.97 (s, 3H), 3.95 (s, 3H); ¹³C NMR (700 MHz, CDCl₃) δ: 193.6, 163.6, 151.9, 149.4, 145.8, 136.3, 129.6, 127.7, 123.7, 120.2, 118.8, 118.7, 117.8, 111.2, 110.3, 56.1 (2); HRMS [M-H]⁺ calculated for C₁₇H₁₆O₄ 283.0970, found 283.0969.

3-(3’,5’-Dimethoxyphenyl)-1-(2-hydroxyphenyl)prop-2-en-1-one (3d):

78.0 % yield, yellow solid; mp. 107–109 °C ¹H NMR (700 MHz, CDCl₃) δ: 12.80 (s, 1H), 7.90 (d, 1H, J = 7.7 Hz), 7.81 (d, 1H, J = 15.4 Hz), 7.58 (d, 1H, J = 15.4 Hz), 7.49 (t, 1H, J = 8.4 Hz), 7.02 (d, 1H, J = 8.4 Hz), 6.93 (t, 1H, J = 7.7 Hz), 6.77 (d, 2H, J = 1.4 Hz), 6.56 (t, 1H, J = 1.4 Hz), 3.83 (s, 6H); ¹³C NMR (700 MHz, CDCl₃) δ: 193.6, 163.5, 161.0, 145.4, 136.4, 136.4, 129.6, 120.5, 119.9, 118.8, 118.6, 106.5, 103.0, 55.4; HRMS [M-H]⁺ calculated for C₁₇H₁₆O₄ 283.0970, found 283.0969.

1-(2-Hydroxyphenyl)-3-(2’,4’,5’-trimethoxyphenyl)prop-2-en-1-one (3e):

90.0 % yield, orange solid; mp. 135–137 °C; ¹H NMR (700 MHz, CDCl₃) δ: 13.08 (s, 1H), 8.21 (d, 1H, J = 15.4 Hz), 7.91 (dd, 1H, J = 7.7, 7.0 Hz), 7.60 (d, 1H, J = 15.4 Hz), 7.46 (dd, 1H, J = 8.4, 7.0 Hz), 6.99 (d, 1H, J = 8.3 Hz), 6.92 (t, 1H, J = 7.9 Hz), 6.5 (s, 1H), 3.94 (s, 3H), 3.92 (s, 3H), 3.90 (s, 3H); ¹³C NMR (700 MHz, CDCl₃) δ: 194.0, 163.5, 155.1, 152.9, 143.2, 140.9, 135.8, 129.5, 120.2, 118.6, 118.4, 117.7, 115.1, 111.7, 96.6, 56.5, 56.2, 56.0; HRMS calculated for C₁₈H₁₈O₅ 314.1154, found 314.1151.

1-(2-Hydroxyphenyl)-3-(3’,4’,5’-trimethoxyphenyl)prop-2-en-1-one (3f):

85.4 % yield, yellow solid; mp. 155–157 °C; ¹H NMR (700 MHz, CDCl₃) δ: 12.86 (s, 1H), 7.92 (d, 1H, J = 7.9 Hz), 7.83 (d, 1H, J = 15.3 Hz), 7.53 (d, 1H, J = 15.3 Hz), 7.49 (t, 1H, J = 7.7 Hz), 7.02 (d, 1H, J = 8.2 Hz), 6.94 (t, 1H, J = 7.5 Hz), 6.87 (s, 2H), 3.93 (s, 6H), 3.91 (s, 3H); ¹³C NMR (700 MHz, CDCl₃) δ: 193.5, 163.6, 153.5, 145.6, 140.8, 136.4, 130.0, 129.6, 120.0, 119.2, 118.8, 118.6, 105.9, 61.0, 56.2; HRMS [M-H]⁺ calculated for C₁₈H₁₈O₅ 313.1076, found 313.1082.

General procedure for the synthesis of methoxy aurones (MA1–6).

Chalcones 3a-f (0.6 mmol) was added to a homogeneous solution of Hg(OAc)₂ (0.221 g, 0.7 mmol) in anhydrous pyridine (20 mL) and the reaction mixture was heated at 110 °C for 12 h. The completion of reaction was marked by consumption of starting material and formation of a single product as visualized by TLC (50 % EtOAc in hexanes). The reaction mixture was then quenched with ice and acidified to the pH of 2 by adding 1.0 N HCl. It was extracted in EtOAc (4 × 50 mL), and the combined extract was washed with water (2 × 50 mL), brine (1 × 50 mL), and dried over anhydrous Na₂SO₄. The drying agent was filtered off and the filtrate was concentrated in vacuo to obtain pure solid products MA1–6. All products were characterized by ¹H NMR, ¹³C NMR and HRMS as follows.

2-(Phenylmethylidene)-2,3-dihydro-1-benzofuran-3-one (MA1):

78.8 % yield, off-white or beige solid; mp. 110–111 °C; ¹H NMR (400 MHz, CDCl₃) δ: 7.92 (d, 2H, J = 7.5 Hz), 7.81 (d, 1H, J = 8.4 Hz), 7.66–7.64 (m, 1H), 7.46 (t, 2H, J = 7.8 Hz), 7.42–7.40 (m, 1H), 7.34 (d, 1H, J = 8.4 Hz), 7.22 (t, 1H, J = 7.4 Hz), 6.90 (s, 1H); ¹³C NMR (700 MHz, CDCl₃) δ: 184.8, 166.2, 146.9, 136.9, 132.3, 131.6, 129.9, 128.9, 124.7, 123.5, 121.6, 113.1, 113.0; HRMS [M-H]⁺ calculated for C₁₅H₁₀O₂ 221.0603, found 221.0596.

2-[(4’-Methoxyphenyl)methylidene]-2,3-dihydro-1-benzofuran-3-one (MA2):

92.5 % yield, yellow solid; mp. 140–142 °C; ¹H NMR (700 MHz, CDCl₃) δ: 7.88 (d, 2H, J = 8.8 Hz), 7.79 (d, 1H, J = 7.6 Hz), 7.64–7.62 (m, 1H), 7.31 (d, 1H, J = 8.3 Hz), 7.20 (t, 1H, J = 7.5 Hz), 6.97 (d, 2H, J = 8.8 Hz), 6.88 (s, 1H), 3.86 (s, 3H); ¹³C NMR (700 MHz, CDCl₃) δ: 184.5, 165.8, 161.1, 145.9, 136.5, 133.4, 125.0, 124.5, 123.3, 121.9, 114.5, 113.4, 112.9, 55.4; HRMS [M-H]⁺ calculated for C₁₆H₁₂O₃ 251.0708, found 251.0701.

2-[(3’,4’-Dimethoxyphenyl)methylidene]-2,3-dihydro-1-benzofuran-3-one (MA3):

quantitative yield, yellow solid; mp. 157–159 °C; ¹H NMR (700 MHz, CDCl₃) δ: 7.81 (d, 1H, J = 7.5 Hz), 7.66–7.63 (m, 1H), 7.54 (d, 1H, J = 1.7 Hz), 7.50 (dd, 1H, J = 8.4, 1.7 Hz), 7.31 (d, 1H, J = 8.3 Hz), 7.22 (t, 1H, J = 7.4 Hz), 6.95 (d, 1H, J = 8.3 Hz), 6.87 (s, 1H), 3.98 (s, 3H), 3.95 (s, 3H); ¹³C NMR (700 MHz, CDCl₃) δ: 184.5, 165.8, 150.9, 149.1, 146.0, 136.6, 126.1, 125.3, 124.6, 123.4, 122.0, 113.8, 113.7, 112.9, 111.3, 56.0 (2). HRMS [M-H]⁻ calculated for C₁₇H₁₄O₄ 281.0814, found 281.0805.

2-[(3’,5’-Dimethoxyphenyl)methylidene]-2,3-dihydro-1-benzofuran-3-one (MA4):

89.0 % yield, yellow solid; mp. 157–160 °C; ¹H NMR (400 MHz, CDCl₃) δ: 7.78 (d, 1H J = 7.2 Hz), 7.66–7.62 (m, 1H), 7.30 (d, 1H, J = 8.3 Hz), 7.20 (t, 1H, J = 7.5 Hz), 7.07 (s, 1H), 7.06 (s, 1H), 6.79 (s, 1H), 6.51 (t, 1H, J = 2.2 Hz), 3.84 (s, 6H); ¹³C NMR (400 MHz, CDCl₃) δ: 184.8, 166.2, 160.9, 147.1, 137.1, 133.9, 124.8, 123.6, 121.7, 113.1, 113.0, 109.6, 102.4, 55.6; HRMS [M+H]⁺ calculated for C₁₇H₁₄O₄ 283.0970, found 283.0979.

2-[(2’,4’,5’-Trimethoxyphenyl)methylidene]-2,3-dihydro-1-benzofuran-3-one (MA5):

99.0 % yield, yellow solid; mp. 240–242 °C; ¹H NMR (700 MHz, CDCl₃) δ: 7.91 (s, 1H), 7.80 (dd, 1H, J = 7.6, 7.0 Hz), 7.63–7.61 (m, 1H), 7.46 (s, 1H), 7.28 (d, 1H, J = 8.2 Hz), 7.20 (t, 1H, J = 7.6 Hz), 6.51 (s, 1H), 3.96 (s, 3H), 3.95 (s, 3H), 3.90 (s, 3H); ¹³C NMR (700 MHz, CDCl₃) δ: 184.3, 165.4, 155.3, 152.4, 145.7, 143.2, 136.2, 124.5, 123.2, 122.2, 114.5, 113.1, 112.8, 108.0, 96.3, 56.6, 56.4, 56.0; HRMS calculated for C₁₈H₁₆O₅ 312.0998, found 312.0998.

2-[(3’,4’,5’-Trimethoxyphenyl)methylidene]-2,3-dihydro-1-benzofuran-3-one (MA6):

89.0 %, yellow solid; mp. 182–183 °C; ¹H NMR (700 MHz, CDCl₃) δ: 7.82 (d, 1H, J = 7.2 Hz), 7.68–7.63 (m, 1H), 7.31 (d, 1H, J = 8.3 Hz), 7.24 (t, 1H, J = 7.4 Hz), 7.19 (s, 2H), 6.84 (s, 1H), 3.95 (s, 6H), 3.93 (s, 3H); ¹³C NMR (700 MHz, CDCl₃) δ: 184.5, 165.9, 153.3, 146.4, 140.1, 136.8, 127.7, 124.7, 123.5, 121.7, 113.4, 112.9, 108.9, 61.0, 56.2; HRMS [M+H]⁺ calculated for C₁₈H₁₆O₅ 313.1076, found 313.1082

General procedure for the synthesis of hydroxy aurones (HA2–6).

The methoxy aurone MA2–6 (0.25 mmol, 1.0 eq) was dissolved in anhydrous CH₂Cl₂ (15 mL) and cooled down to 0 °C. BBr₃ (1 mmol, 4.0 eq) was added slowly to the reaction mixture under N₂ atmosphere and stirred. The reaction mixture was allowed to attain room temperature and stirring continued for 12 h. TLC examination (50 % EtOAc in hexanes) revealed the completion of the reaction. The reaction mixture was then cooled to 0 °C and carefully quenched with slow drop-wise addition of water until the excess BBr₃ reacted completely. The precipitated solid product was filtered, washed with water, and dried over CaCl₂ in a vacuum desiccator. The crude product thus obtained was purified by column chromatography over Si gel using 10 % MeOH in CH₂Cl₂ to afford pure hydroxyl aurones HA2–6. All hydroxy aurones were characterized by ¹H NMR, ¹³C NMR and HRMS as follows.

2-[(4’-Hydroxyphenyl)methylidene]-2,3-dihydro-1-benzofuran-3-one (HA2):

89.8% yield, yellow solid; mp. 264–266 °C; ¹H NMR (700 MHz, DMSO-d₆) δ: 10.23 (s, 1H), 7.88 (d, 2H, J = 8.4 Hz), 7.79–7.77 (m, 2H), 7.54 (d, 1H, J = 8.6 Hz), 7.30 (t, 1H, J = 7.4 Hz), 6.91 (d, 3H, J = 8.8 Hz); ¹³C NMR (700 MHz, DMSO-d₆) δ: 183.2, 165.0, 159.8, 144.7, 137.2, 133.8, 124.1, 123.7, 122.9, 121.3, 116.2, 113.5, 113.2; HRMS [M+H]⁺ calculated for C₁₅H₁₀O₃ 239.0708, found 239.0718.

2-[(3’,4’-Dihydroxyphenyl)methylidene]-2,3-dihydro-1-benzofuran-3-one (HA3):

92.0 % yield, yellow solid; mp. 231–233 °C; ¹H NMR (400 MHz, DMSO-d₆) δ: 9.82 (bs, 1H), 9.34 (bs, 1H), 7.80–7.76 (m, 2H), 7.53–7.50 (m, 2H), 7.34–7.28 (m, 2H), 6.86 (d, 1H), 6.82 (s, 1H); ¹³C NMR (700 MHz, DMSO-d₆) δ: 183.1, 165.0, 148.6, 145.7, 144.7, 137.1, 125.2, 124.1, 123.7, 123.3, 121.4, 118.3 (d), 116.1 (d), 114.0 (d), 113.0; HRMS [M+H]⁺ calculated for C₁₅H₁₀O₄ 255.0657, found 255.0660.

2-[(3’,5’-Dihydroxyphenyl)methylidene]-2,3-dihydro-1-benzofuran-3-one (HA4):

80.0 % yield, grey solid; decomposed at 250 °C; ¹H NMR (700 MHz, DMSO-d₆) δ: 7.79 (t, 2H, J = 7.4 Hz), 7.50 (d, 1H, J = 8.6 Hz), 7.30 (t, 1H, J = 7.3 Hz), 6.88 (d, 2H, J = 2.3 Hz), 6.73 (s, 1H), 6.36 (t, 1H, J = 2.0 Hz); ¹³C NMR (700 MHz, DMSO-d₆) δ: 183.7, 165.4, 158.7, 146.2, 137.8, 133.2, 124.4, 124.0, 121.0, 113.1, 113.1, 109.7, 105.0; HRMS [M-H]⁻ calculated for C₁₅H₁₀O₄ 253.0501, found 253.0513.

2-[(2’,4’,5’-Trihydroxyphenyl)methylidene]-2,3-dihydro-1-benzofuran-3-one (HA5):

83.0 % yield, red solid; mp. 191–193 °C; ¹H NMR (400 MHz, Acetone-d₆) δ: 8.94 (s, 1H), 8.71 (s, 1H), 7.97 (s, 1H), 7.85 (s, 1H), 7.71–7.76 (m, 2H), 7.43 (d, 1H, J = 8.2 Hz), 7.40 (s, 1H), 7.28 (t, 1H, J = 7.4 Hz), 6.57 (s, 1H); ¹³C NMR (400 MHz, Acetone-d₆) δ: 183.9, 166.1, 153.8, 150.7, 145.6, 139.7, 137.2, 124.7, 124.1, 123.0, 117.7, 113.7, 111.6, 108.9, 103.7; HRMS calculated for C₁₅H₁₀O₅ 270.0528, found 270.0529.

2-[(3’,4’,5’-Trihydroxyphenyl)methylidene]-2,3-dihydro-1-benzofuran-3-one (HA6):

86.0 % yield, greenish-yellow solid; decomposed at 256 °C; ¹H NMR (700 MHz, DMSO-d₆) δ: 9.28 (s, 2H), 9.05 (s, 1H), 7.78–7.76 (m, 2H), 7.48 (d, 1H, J = 8.5 Hz), 7.28 (dd, 1H, J = 14.8, 7.3 Hz), 7.03 (s, 2H), 6.72 (s, 1H); ¹³C NMR (700 MHz, DMSO-d₆) δ: 183.0, 164.9, 146.2, 144.8, 137.1, 137.0, 124.2, 123.7, 122.0, 121.4, 114.5, 113.0, 111.3; HRMS [M-H]⁻ calculated for C₁₅H₁₀O₅ 269.0450, found 269.0445.

Biofilm inhibition assays.

Biofilm inhibition assays were performed in polystyrene microtiter 96 well plates. Stock solutions were prepared in chemically defined medium (CDM, JRH Biosciences, Lenexa, KS) with 2 % sucrose, 1 % bacteria cultures and various concentrations of the small molecule inhibitors to examine their activity against biofilm formation as described^{69, 70}. These stocks were assayed in 96 well plates in triplicate and incubated at 37 °C and 5 % CO₂ for 16 h. After reading optical density for bacterial growth, the plate was gently washed with water, dried, and stained with crystal violet, and then gently rinsed again with deionized water leaving the stained biofilm at the bottoms of the wells. Biofilms was dissolved in 200 μL of 30 % acetic acid and absorbance at 562 nm was used read to determine biofilm biomass. Each assay was carried out at least in triplicate. Biofilm inhibitory concentration (IC₅₀) of the compounds was determined by serial dilutions.

Gtf inhibition determined by glucan quantification assays.

Overnight cultures of S. mutans UA159 were centrifuged (6500rpm, 4 °C, 10min) to remove cells. Supernatant was mixed with ethanol (1:1) and incubated at −80 °C for 1h. The precipitated Gtfs were palleted using centrifugation and resuspended in chemically defined media (CDM) 10 μL of Gtfs suspended in CDM were assayed on Ibidi slides with varying concentrations of inhibitor, 1 % sucrose, 1 % DMSO and 1uM Cascade blue dye in CDM. The slides were then incubated at 37 °C with 5 % CO₂ for 16 h after which, the wells of Ibidi slides were gently rinsed with 1x PBS and treated with 1x PBS for fluorescence microscopy imaging. The images obtained were processed in ImageJ to quantify glucans and graphed in GraphPad Prism.

S. mutans, S. gordonii, and S. sanguinis growth assays.

Effects of compounds on S. mutans and commensal bacterial growth were evaluated using the growth assay as described⁶⁹. S. mutans UA159, S. gordonii DL1, S. sanguinis SK36, cultures were grown for 24 h under 5 % CO₂ at 37 °C. These cultures were then reinoculated with fresh THB (1:5) until OD₄₇₀ = 1 when the bacteria were ready to be used. Different concentrations of the inhibitor were assayed in chemically defined media (CDM) with 1 % of the bacteria, 1 % sucrose and 1 % DMSO in 96 well plates. The 96 well plates were incubated under 5 % CO₂ at 37 °C for 16 h. Growth of the bacteria was read after 16 h at OD₄₇₀. Each assay was carried out at least in triplicate.

Synthesis of porous cubic manganese oxide microparticle templates.

Porous Mn₂O₃ microparticle templates of 3 μm in size were synthesized as described previously^{63, 65}. Briefly, a nano-seed solution was prepared by mixing 0.04 g of NH₄HCO₃ and 0.02 g of MnSO₄ in DI water (200 mL). Then, the nano-seed solution (80 mL) was added to a 6 mM of MnSO₄ (1000 mL) followed by 6 mM solution of NH₄HCO₃ (1000 mL) both containing 2-propanol (0.5 % vol) was added to the nano-seed solution and was heated at 60 °C for 30 minutes to produce 3 μm cubic manganese carbonate particles. Once collected and dried via filtration, the Mn₂CO₃ microparticles were heated at 650°C for 3.5 h in the muffled oven to produce porous Mn₂O₃ microparticles.

Synthesis of cubical hydrogel microparticles.

pH-Responsive cubic hydrogel cubic microparticles were synthesized by depositing hydrogen-bonded [PMAA/PVPON]n (the subscript denotes the number of polymer bilayers) multilayers at the surfaces of Mn₂O₃ microparticle templates. The porous templates were first exposed to an aqueous poly(ethyleneimine) (PEI) solution in deionized (DI) water (1.5 mg/mL) for 1 h to enhance the adsorption of the following (PMAA/PVPON) layers to the particle surfaces followed by deposition of the polymers from aqueous polymeric solutions (1.5 mg/mL) at pH = 2 for 45 min each. The polymer deposition was achieved through sonication (15 min) and shaking (30 min) of the manganese oxide porous templates in polymer solutions. After the deposition of each layer, the template particles were centrifuged for 10 min at 4,900 rpm and re-suspended in phosphate buffer solution (0.01 M, pH = 2) twice to rinse away excess polymer before the next deposition cycle. Following the deposition of a 5-bilayer (PMAA/PVPON)₅ coating, the PMAA layers were cross-linked with ethylenediamine by, first, activating the PMAA carboxylic groups with a carbodiimide solution (5 mg/mL, pH = 5, 0.01 M phosphate) for 30 min, then exposing the particles to ethylenediamine (12 μL/mL in 0.01 M phosphate, pH = 5.8) for 16 h. Afterwards, PVPON was removed from the PMAA network by exposing the core-shell particles to 0.01 M phosphate buffer solution (pH = 8.5) for 24 h while shaking. Cubic PMAA hydrogel microparticles were obtained after dissolving the manganese oxide core in hydrochloric acid solution (8M HCl) for 24 h. The hydrogel microparticles were treated with ethylenediamine tetraacetic acid disodium salt solution (EDTA, 0.1 M) at pH = 7 overnight by shaking to remove any residual manganese ions in the hydrogel network. The PMAA hydrogel microparticles were then purified by dialysis in DI water for 3 days using a Float-a-Lyzer (Fisher; MWCO 20 kDa).

Rat model of dental caries.

In vivo studies of colonization and virulence of S. mutans were evaluated using a previously reported rat model of dental caries⁷¹. Offspring of gnotobiotic Fischer 344 rats used in this experiment were bred and maintained in trexler isolators. Male and female rat pups were removed from isolators at 20 days of age and randomly assigned into 5 treatment groups of 5 rats / group in cages with filter tops. Rats were then infected with S. mutans UA159 strain by oral swabbing daily for four consecutive days with a fresh overnight culture of S. mutans UA159. Rats were provided with caries promoting Teklad Diet 305 containing 5 % sucrose (Harlan Laboratories, Inc., Indianapolis, IN) and sterile drinking water ad libitum. Oral swabs were taken 5 days post-infection and plated on Todd Hewitt (TH) agar plates and incubated at 37 °C in an environment of 5 % CO₂ in the air to confirm colonization. Rats were weighed at weaning and at the termination of the experiment. One-week post-infection, the molars of the rats were treated topically twice daily for 4 weeks with the test compounds using camel-hair brushes. The five treatment groups used in this study were: 1) HEBI (100 μM); 2) HA5 (100 μM); 3) hydrogel encapsulated PBS (no drug) containing 0.1 % DMSO (negative control), 4) 250 ppm NaF (positive control) and 5) infected untreated group (negative control). Drinking water was withheld for 60 min following each treatment with the compound. Animals were weighed at weaning and at the termination of the experiment. On day 60, the rats were sacrificed using CO₂ followed by cervical dislocation or bilateral thoracotomy. The mandibles were surgically removed and cleaned of excess tissue to assess the level of bacteria present and the extent of caries formation. The right mandibles from each rat were placed in a tube containing phosphate buffer (3 mL), placed on ice and sonicated (10 sec) to release bacteria from the molars. Each sample was serially diluted, plated on blood agar plates (BAP) and mitis-salivarius (MS) agar plates and incubated in an environment of 5 % CO₂ at 37 °C to quantify the level of total bacteria and S. mutans present in the plaque. The right and left mandibles from each rat were then placed in 95 % ethanol for 24 h. The mandibles will be cleaned and stained overnight with murexide solution. After drying, the mandibles were sectioned and scored for caries activity using the Keyes method⁶⁸. Caries scores were recorded for the buccal, sulcal and proximal molar surfaces individually so that differences among the surfaces can be distinguished. Statistical significance in the mean caries scores, colony-forming units (CFUs) / mandible and body weights between groups of rats were determined by one-way analysis of variance (ANOVA) with the Tukey-Kramer multiple comparison test using the InStat program (Graphpad Software, San Diego, CA). When determining the statistical significance between the two groups, an unpaired t-test was applied. Differences between groups were considered significant at a P-value < 0.05. All experimental protocols were approved by the University of Alabama at Birmingham Institutional Animal Care and Use Committee (protocol No: IACUC-20047). The methods were carried out in accordance with the relevant guidelines and regulations.

ABBREVIATIONS.

AFM

Atomic Force Microscope

BOC

tert-Butyloxy carbonyl

BrS

Broad signal

CDM

Chemically defined medium

CFU

Colony forming unit

DMAP, N

N-Dimethyl aminopyridine

DMSO

Dimethyl sulfoxide

EDC

N-Ethyl-N’-(3-dimethylaminopropyl)carbodiimide

EPS

Extracellular polysaccharide

Gtf

Glucosyltransferases

HEBI

Hydrogel encapsulated biofilm inhibitor

HRMS

High resolution mass spectrometry

IC₅₀

The half maximal inhibitory concentration

LC-MS

Liquid-chromatography mass spectrometry

NMR

Nuclear magnetic resonance

SAR

Structure activity relationship

SDS-PAGE

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis

S. gordonii

Streptococcus gordonii

S. mutans

Streptococcus mutans

S. sanguinis

Streptococcus sanguinis

TFA

Trifluoroacetic acid

THB

Todd Hewitt Broth