Caffeic acid phenethyl ester inhibits multispecies biofilm formation and cariogenicity

Bacterial culture

S. mutans (ATCC 25175), S. oralis (ATCC 35037) and S. mitis (ATCC 49456T) were used in this study. Mitis salivarius bacitracin agar (HiMedia Technology, Shenzhen, China) was used as the selective media for S. mutans and mitis salivarius agar (HiMedia Technology, China) served as the selective media for S. oralis and S. mitis. The bacteria were separately or co-inoculated in brain heart infusion broth (BHI, Becton, Dickinson, Franklin Lakes, NJ, USA) overnight at 37 °C under aerobic conditions with 5% CO₂ before performing the experiments.

Bacterial susceptibility test

Bacteria were maintained in BHI broth overnight, and the bacterial number was assessed at an optical density (OD) of 600 nm. The bacterial suspension of S. mutans, S. oralis or S. mitis was adjusted to 2 × 10⁷ CFU/ml and transferred to a 96-well plate (100 µl/well). CAPE was purchased from Sigma-Aldrich (St. Louis, MO, USA). A stock solution of CAPE was prepared in 50% dimethyl sulfoxide (DMSO; Sigma-Aldrich, St. Louis, MO, USA)-phosphate buffered saline (pH 7.4). To test bacterial susceptibility, CAPE at stock concentration was diluted in BHI media to final concentrations of 0.8, 1.6, 3.2 or 6.4 mg/ml in 5% DMSO in each well, based on a previous study (Niu et al., 2020). Bacteria cultured in BHI, 5% DMSO-BHI and 50 µg/ml chlorhexidine (CHX; Thermo Fisher Scientific, Waltham, MA USA) served as growth, solvent and positive controls, respectively. The plate was incubated at 37 °C for 48 h. The plate was measured for absorbance at 600 nm in a MULTISKAN Sky microplate reader (Thermo Fisher Scientific, Waltham, MA, USA). The OD 600 nm was converted to percent bacterial growth normalized to the solvent control. To determine the concentration of CAPE needed to inhibit bacterial growth by 50%, the 50% inhibitory concentration (IC₅₀) of CAPE against the bacteria was analyzed. The IC₅₀ was calculated from the concentration-response curve. Three independent experiments were performed.

Multispecies biofilm formation study

Dental biofilm forms on the hydroxyapatite (HA)-containing enamel surface. To mimic the dental surface for biofilm adhesion, HA-coated plates were prepared by modifying the protocol in a previous study (Schilling et al., 1994). Briefly, a calcifying solution containing 2.5 mM CaCl₂·2H₂O, 7.5 mM KH₂PO₄ and 250.0 mM triethanolamine at pH 7.4 was prepared and placed in 96-well microplates (Corning, Oneonta, NY, USA). The plates were incubated at 75 °C for 3.5 h for two cycles. The plates were dried at room temperature and sterilized with ethylene oxide. Bovine serum albumin (1%) in phosphate buffer was added in the HA-coated microplates and incubated for 1 h at 37 °C.

The multispecies bacteria comprising S. mitis, S. oralis and S. mutans was mixed at a 1:1:1 ratio in BHI broth supplemented with 1% sucrose and placed in the HA-coated wells. CAPE dilutions were added to each well at final concentrations of 0.8, 1.6 and 3.2 mg/ml. Bacteria grown in BHI, 5% DMSO-BHI and 50 µg/ml CHX served as growth, solvent and positive controls, respectively. The plate was incubated at 37 °C for 48 h. Non-adherent bacteria were washed out with phosphate buffered saline. Multispecies biofilm adhered on HA-coated wells was fixed with absolute methanol for 15 min and stained with 0.1% crystal violet for 5 min. The stained biofilm was solubilized with 33% glacial acetic acid. The OD at 592 nm was determined using a microplate reader. Biofilm formation was reported as a percentage normalized to the solvent control. Three independent experiments were conducted.

Total RNA extraction and cDNA synthesis

S. mitis, S. oralis and S. mutans co-culture was performed at a 1:1:1 ratio in BHI broth supplemented with 1% sucrose. The bacterial co-culture was treated with 1.6 mg/ml CAPE at 37 °C for 24 h. Bacteria treated with 5% DMSO served as the solvent control. Total RNA was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA). Total RNA concentration and purity were determined by a NanoDrop 2000 (Thermo Fisher, Waltham, MA, USA) with a 260/280 ratio in a range of 1.8–1.9. Total RNA (500 ng) was used to synthesize cDNA using a PrimeScript 1st strand cDNA Synthesis Kit (Takara Bio Inc., Shiga, Japan). The total volume of the reverse transcription (RT) mixture was 10 µl: total RNA, 6.5 µl; 5× primer Script Buffer, 2 µl; PrimeScript RT Enzyme Mix, 0.5 µl; Oligo dT Primer, 0.5 µl; and Random 6 mers, 0.5 µl. The RT conditions were: RT at 37 °C for 15 min, inactivation of reverse transcriptase at 85 °C for 5 s and hold at 4 °C. The cDNA was stored at −20 °C until use.

Real-time quantitative PCR (qPCR)

The PCR primers used in this study were from a previous study (Wen et al., 2010) (Table 1). 16s rRNA gene was used as an internal control (Thanetchaloempong, Koontongkaew & Utispan, 2022). To test the specificity of the primers, SYBR green based real time PCR was performed in the QuantStudio™ 3 Real‑Time PCR System (Thermo Fisher Scientific, Waltham, MA USA). The specific real time PCR condition was selected when the specific melting curve of the PCR product resulted in the tested primer group, while a disappearing melting curve/ cycle threshold (CT) was detected in the non-template control group.

Real-time qPCR was performed using a KAPA SYBR® FAST qPCR Kit Master Mix (Kapa Biosystems, Wilmington, MA, USA). The total volume of the qPCR system was 20 µl: cDNA (10 ng/µl), 2 µl; sense/antisense primer (10 µM each), 0.4 µl each; SYBR Green Master Mix (Kapa Biosystems, Wilmington, MA, USA), 10 µl; and ddH₂O, 7.2 µl. qPCR was performed in a two-step method in the QuantStudio™ 3 Real‑Time PCR System (Thermo Fisher Scientific, Waltham, MA USA). The qPCR conditions were: initial denaturation at 95 °C for 10 min, followed by 40 cycles of denaturation at 95 °C for 15 s and annealing at 60 °C for 60 s. The melting curve analysis was performed at the end of the amplification from 60 °C to 95 °C and with a hold of 1 s every 0.16 °C. The original CT values of all genes were obtained from QuantStudio™ Design & Analysis Software (Thermo Fisher Scientific, Waltham, MA USA). The relative gene expression was evaluated by the difference in cycle threshold (ΔCT) between the CAPE-treated multispecies bacteria and the solvent control using the 2^−ΔΔCT equation [ΔCT = CT (CAPE treatment) − CT (solvent control)] (Rao et al., 2013). Three independent experiments were conducted.

Glucosyltransferase activity assay

Multispecies or monospecies bacteria-derived glucosyltransferase (GTF) was prepared by modifying the protocol from a previous study (Linzer & Slade, 1976). Briefly, single bacteria (1 × 10⁸ CFU/ml) or mixed S. mitis, S. oralis and S. mutans (1:1:1 ratio) were cultured in sucrose-free tryptic soy broth at 37 °C for 48 h. The cultured suspension was centrifuged at 10,000 × g for 30 min, and the supernatant was collected. Crude GTF was precipitated with 45% ammonium sulphate at 4 °C for 48 h. The precipitated enzyme was dissolved in 15 ml phosphate buffer, pH 7.4 and dialyzed for 4 days at 4 °C against 0.05 M phosphate buffer, pH 6.8 (4 L). The buffer was changed every 24 h. The dialyzed enzyme was lyophilized for 48 h. The enzyme powders were dissolved in 0.05 M phosphate buffer, pH 7.4. The protein concentration was determined using a Pierce^TM BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA USA) according to the manufacturer’s directions. The enzyme was kept at 80 °C and diluted in phosphate buffer to 200 mg/ml prior to use.

To determine GTF activity, 50 µl of the enzyme (200 mg/ml), was mixed with CAPE at final concentrations of 0.8, 1.6 and 3.2 mg/ml in sucrose (30 µmole) containing 0.6 M acetate buffer pH 5.5 to a final volume of 300 µl. The buffer without CAPE was used as control. The reaction mixture was incubated at 37 °C for 3 h and further heated at 100 °C for 5 min. The enzyme was heated to 100 °C for 5 min. The glucans produced from GTF activity were precipitated by centrifugation at 20,000 × g for 6 min. The glucans were collected by treating them with a 5% phenol-containing sulphuric acid solution, heated to 110 °C for 5 min and cooled to room temperature. Glucan production was measured via the OD at 490 nm in a microplate reader (Thermo Fisher Scientific, Waltham, MA, USA). The GTF activity was expressed as the percentage of glucan production compared with the control group using the formula [(OD treatment − OD blank)/(OD control − OD blank)] × 100. Three independent experiments were conducted.

Confocal laser scanning microscopy

HA-coated surfaces were formed in eight-well chamber glass slides (SPL Lifesciences Co. Ltd, Gyeonggi-do, Korea) as previously described. The multispecies bacteria comprising S. mitis, S. oralis and S. mutans (1:1:1 ratio) were co-cultured in BHI supplemented with 1% sucrose in the HA-coated slides and incubated at 37 °C for 24 h. Non-adherent bacteria were removed from the well and washed with phosphate buffer. CAPE (0.8 and 1.6 mg/ml) was added in each well and incubated at 37 °C for 24 h. The biofilm in 5% DMSO-BHI and 50 µg/ml CHX served as solvent and positive controls, respectively. The biofilm was stained using the Live/Dead BacLight Bacterial Viability kit (Invitrogen Molecular Probes, Carlsbad, CA, USA). Live and dead bacteria were visualized as green and red signals, respectively, using a confocal laser scanning microscope (CLSM) (FLUOVIEW FV3000 Olympus, Tokyo, Japan). To analyze the bacterial viability within the biofilm and total biofilm thickness, the Z-stack images of the stained biofilm were obtained using the cellSens Dimension software (Olympus, Tokyo, Japan). Three independent experiments were conducted.

Statistical analysis

The data are quantitative and presented as the means and standard error of the mean (SEM). The unpaired t-test was performed to compare the gene expression level between the solvent control and the CAPE-treated group. For multiple group comparison of percent bacterial growth, percent GTF activity, live/dead bacterial ratio and biofilm thickness, one-way ANOVA was applied, followed by Dunnett’s post hoc test using Prism GraphPad 8.0 (GraphPad Software, La Jolla, CA, USA). Significance was determined at p ≤ 0.05.

Antibacterial effect of CAPE

To confirm the antibacterial effect of CAPE, the cytotoxic effect of the CAPE solvent (5% DMSO) was evaluated on S. mitis, S. oralis and S. mutans. The solvent control maintained bacterial growth similar to the control group (p > 0.05), thus, it was non-toxic to the tested bacteria (Fig. 1). Therefore, the solvent control was used to compare the CAPE effects in subsequent experiments. CAPE at 0.8, 1.6, 3.2 and 6.4 mg/ml exhibited a significant antibacterial effect on S. mitis (Fig. 1A) and S. mutans (Fig. 1C) compared with the solvent control (p < 0.05). CAPE at 1.6, 3.2 and 6.4 mg/ml significantly decreased the percent bacterial growth in S. oralis compared with the solvent control (p < 0.05) (Fig. 1B). The 50% inhibitory concentration (IC₅₀) value of CAPE was also calculated. The IC₅₀ of CAPE (mean ± SEM) against S. mitis, S. oralis and S. mutans was 0.87 ± 0.02, 1.39 ± 0.09 and 1.66 ± 0.14 mg/ml, respectively. Based on the IC₅₀ data, S. mutans had the highest resistance to CAPE among the tested bacteria.

CAPE affected adhesion and cariogenic gene expression in the multispecies biofilm

A multispecies biofilm comprised of S. mitis, S. oralis and S. mutans (1:1:1) was established. The effect of CAPE on the adhesion and gene expression of virulence factors in the multispecies biofilm were investigated. The results demonstrated that 0.8, 1.6 and 3.2 mg/ml CAPE significantly reduced multispecies biofilm formation by 0.5–70% compared with the solvent control (p < 0.05) (Fig. 2A). Moreover, 1.6 mg/ml CAPE significantly decreased gene expression involved in sucrose-dependent adhesion (gtfB and gbpB), acidogenicity (ldh), aciduricity (brpA) and quorum sensing (luxS) in multispecies biofilm compared with the solvent control (p < 0.05) (Fig. 2B). However, there was no significant difference in sucrose-independent adhesion (spaP) gene expression in the multispecies biofilm between the CAPE treatment and solvent control group (p > 0.05).

CAPE reduced GTF activity in monospecies and multispecies bacterial cultures

The intrinsic GTF activity in monospecies and multispecies bacterial cultures is presented in Table 2. S. oralis and S. mutans expressed relatively high GTF activity compared with that of S. mitis. Moreover, the maximum GTF activity was found in the multispecies bacterial culture model. Therefore, the changes in GTF activity after CAPE treatment in the monoculture of S. oralis and S. mutans and multisspecies co-culture comprising S. mitis, S. oralis and S. mutans were evaluated. The results indicated that 0.8 and 1.6 mg/ml CAPE significantly decreased GTF activity in S. oralis compared with control (p < 0.05) (Fig. 3A). However, 3.2 mg/ml CAPE did not significantly change GTF activity in S. oralis. Interestingly, CAPE at all tested concentrations significantly inhibited GTF activity in S. mutans to approximately 48% compared with control (p < 0.05) (Fig. 3B). In addition, CAPE significantly reduced GTF activity in the multispecies co-culture system in a dose-dependent manner compared with control (p < 0.05) (Fig. 3C).

Bactericidal penetrative effect of CAPE in multispecies biofilm

The bactericidal penetrative effect of 0.8 and 1.6 mg/ml CAPE was assessed in the multispecies biofilm using CLSM. Z-stack images displayed the bactericidal depth of CAPE from the outermost into innermost layers of the multispecies biofilm (Fig. 4A). We found that 0.8 and 1.6 mg/ml CAPE significantly decreased the live/dead bacterial ratio in the multispecies biofilm compared with the solvent control (p < 0.05) (Fig. 4B).

Effect of CAPE on the reduction of the multispecies biofilm thickness

The thickness of the 0.8 and 1.6 mg/ml CAPE-treated multispecies biofilm was evaluated using CLSM. The confocal images exhibited 3D structures with live/dead bacteria in the multispecies biofilm at all conditions (Fig. 5A). Measurement of the multispecies biofilm structure indicated that 1.6 mg/ml CAPE significantly reduced the biofilm thickness compared with the solvent control (p < 0.05) (Fig. 5B).