Real-Time Monitoring of Biofilm Formation Using a Noninvasive Impedance-Based Method

Abstract

Biofilms are complex three-dimensional microbial communities that adhere to a variety of surfaces and interact with their surroundings. Because of the dynamic nature of biofilm formation, establishing a uniform technique for quantifying and monitoring biofilm volume, shape, and features in real-time is challenging. Herein, we describe a noninvasive electrochemical impedance approach for real-time monitoring of dental plaque-derived multispecies biofilm growth on a range of substrates. A working equation relating electrochemical impedance to live biofilm volume has been developed that is applicable to all three surfaces examined, including glass, dental filling resin, and Ca²⁺-releasing resin composites. Impedance changes of 2.5, 35, 50, and 65% correlated to biofilm volumes of 0.10 ± 0.01, 16.9 ± 2.2, 29.7 ± 2.3, and 38.6 ± 2.8 μm³/μm², respectively. We discovered that glass, dental filling resin, and Ca²⁺-releasing dental composites required approximately 3.5, 4.5, and 6 days, respectively, to achieve a 50% change in impedance. The local pH change at the biofilm-substrate interfaces also monitored with potentiometry pH microsensor, and pH change varied according to biofilm volume. This impedance-based technique can be a useful analytical method for monitoring the growth of biofilms on a variety of substrates in real-time. Therefore, this technique may be beneficial for examining antibacterial properties of novel biomaterials.

1. INTRODUCTION

Biofilms are ubiquitous and formed by microorganisms when they attach to a surface and subsequently proliferate. Biofilms are found in both natural and man-made environments and are involved in a wide range of biological processes, including water and wastewater treatment, corrosion, medical device infections, and food spoilage [1–4]. Due to their complex three-dimensional structure, biofilms are more resistant to antimicrobial treatment than are individual bacteria [5]. Thus, it is essential to understand real-time biofilm development and its metabolism for effective antibacterial strategies. The formation of biofilms is a complex, dynamic process that can involve a variety of microorganisms. The first step in the biofilm cycle is the attachment of microorganisms to a surface. This initial attachment is followed by the proliferation of the attached cells and the organization of the cells into a biofilm through the production of an extracellular matrix. The final stage of biofilm development is the maturation of the biofilm, which can lead to the formation of a stable biofilm community [3,6].

The development of biofilms can be monitored with a variety of methods, including confocal laser scanning microscopy (CLSM), scanning electron microscopy (SEM), and atomic force microscopy (AFM) [7–10]. CLSM has been used to visualize and quantify the spatial structure of the biofilm. However, bleaching of dyes during experiment can put CLSM to the disadvantage for real-time measurement. There is no known study of biofilm formation monitoring using CLSM in real time continuously for 7 days. The conventional SEM technique has been used extensively to study the high-resolution spatial structure of biofilms, biofilm formation kinetics, and the effect of antibiotics on biofilm. AFM has also been used for the qualitative and quantitative assessment of living bacteria under physiological conditions and for in situ imaging of bacteria topography. Significant limitations of the AFM technique are the inability to perform large area survey scans and the presence of artifacts caused by tip shape and size. Overall, microscopic techniques require sample preparation such as fixation, dehydration and conducting material coating that change physiological state of biofilm [7,9,11,12]. These methods might not be able to continuously monitor the slow biofilm development which grew for 5–7 days frame on various substrates. Also, collection of data suffered from user’s unintended bias in collecting images in different parts of the biofilm.

Electrochemical techniques such as amperometry, potentiometry, and impedimetry have been extensively reported in the study of biofilms [13,14]. In previous work by our group, we demonstrated the ability and versatility of amperometric H₂O₂ and potentiometric pH sensors for biofilm studies [15–18]. Moreover, electrochemical impedance spectroscopy (EIS) has been a useful and non-destructive technique to study for various applications including thin films deposition and characterization, assessment of the biofilm corrosion, formation of polymeric films, impedance probing of biological tissues, and the biosensor’s surface characterization [19–22]. Additionally, the impedance-based method has been widely used to detect bacteria and quantify biofilm because of its high sensitivity to bacterial attachment and metabolites released during biofilm growth [23–25]. Kim et al. used electrochemical impedance spectroscopy to detect the early adhesion and maturation of a Pseudomonas aeruginosa bacterial biofilm [26]. They were able to observe a decrease in double-layer capacitance at specific points in time up to 72 hours (0, 6, 24, 48, 72 h); however, there was no real-time growth rate reported as a continuous measurement from 0 to 72 h. In addition, with the same bacteria, the Brönstrup group used single-frequency electrochemical impedance spectroscopy to characterize growth and antibiotic effects [27]. They created a calibration curve with the cell index and impedance measurement so that they could indicate decreased or increased growth rate. To the best of our knowledge, all the reported studies were performed on the printed electrode surfaces, which makes them unsuitable to be used to monitor the biofilm growth on different substrates or new biomaterials [28–30].

Herein, we report the development of a novel impedance-based approach for monitoring the production of biofilms on a variety of biomaterials in real-time. In addition, we used our previously developed pH sensors to monitor the metabolic activities of the bacteria in terms of acid production as they adsorb and grow on the various substrates. Combining biofilm growth from impedance-based method and metabolic activity data from pH sensor is a critical tool in comprehending biofilm formation and developing strategies for biofilm control. In this study, we aimed to demonstrate that the impedance-based method can be used to effectively monitor biofilm formation on a variety of biomaterials in a real-time, noninvasive manner.

2. MATERIALS AND METHODS

2.1. Materials.

D-glucose was purchased from Macron. Protease peptone was purchased from Hardy Diagnostics, and Difco tryptone (pancreatic digest of casein) from Life Technologies Corp. Carboxymethylcellulose sodium salt and poly(styrene sulfonic acid) sodium salt, (PSS M.W. 70,000) were purchased from Alfa Aesar. 3,4-ethylenedioxythiophene (EDOT, 97%), dichloromethane and yeast extract were obtained from Sigma Aldrich. Ferrocene methanol was purchased from TCI America. A 125 μm Teflon-coated platinum (Pt)/iridium wire was obtained from World Precision Instruments.

Baseline mucin medium (BMM) was prepared by mixing 10 g of protease peptone, 5 g of Difco tryptone (pancreatic digest of casein), 5 g of yeast extract, 8 g of the sodium salt of carboxymethylcellulose sodium salt, and 2.5 g of KCl in 1.0 L of deionized water, followed by autoclaving for 15 min at 120°C.

2.2. Instrumentation.

All electrochemical experiments were performed with a CHI electrochemical workstation (model no. 760E, CH Instruments, Austin, TX, USA). A CHI multiplexer (CHI 684) was used for the multichannel setup for the impedance measurements. All impedance measurements were performed in two-electrode configurations. Potentiometric measurement was performed by using a high impedance potentiometer (model no. 9101, Lawson Laboratories, Malvern, PA, USA). All fluorescence imaging was investigated with confocal microscopy (model no. LSM 780 NLO, Zeiss, Oberkochen, Germany).

2.3. Fabrication and characterization of pH and impedance sensors.

The impedance wire sensor was fabricated from the 125 μm Teflon-coated Pt wire. Approximately 1.5 mm of the Teflon coating was removed, and the Pt wire washed with absolute alcohol. Later, poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT-PSS) was electrodeposited through cyclic voltammetry from −0.2 to 1.0 V vs Ag/AgCl on the exposed Pt wire by using ~10 mM EDOT and 0.1 mM PSS in an aqueous solution in a three-electrode configuration. Fabrication of pH sensor and characterization were previously described and discussed [16].

Impedance sensors were tested before all biofilm-related experiments. Full-spectrum impedance was recorded in a three-electrode configuration by using the CHI workstation (model no. 760E) in 10 mM phosphate-buffered solution (PBS) (pH 7.2). The impedance sensor was used as the working electrode (WE), Ag/AgCl (1M KCl) as the RE, and Pt wire as the CE. Sensors were characterized at 0.1 V (double-layer potential was decided after running cyclic voltammetry, CV, at a 50 mV/s scan rate) in the frequency range from 100 kHz to 0.1 Hz at 5 mV amplitude. For real-time monitoring of biofilm growth, single-frequency impedance (1 Hz) was recorded at 0.1 V with a sinusoidal amplitude of 5 mV in a two-electrode configuration. PEDOT-PSS-deposited sensors were used as the WE and the RE. A CHI multiplexer (CHI 684) with the CHI workstation (model no. 760E) was used for the independent multiple channel operation. Each sensor was monitored every ~20 min with the same electrical conditions during the experiment. In this study, we used the normalized impedance over time procedure developed by Farrow et al. (2012) [31].

$$\%\mathit{\text{ change in Impedance}}=\frac{Z_{T,1Hz-}{Z}_{BMM,1Hz}}{Z_{PBS,1Hz}}\times 100=\frac{\Delta Z_{T,1Hz}}{Z_{PBS,1Hz}}\times 100$$

The relative change in total impedance (Z_T,1Hz – Z_BMM,1Hz) of biofilm growth was calculated by subtracting stable background impedance signals in BMM (Z_BMM,1Hz) from real-time impedance signals for biofilm growth (Z_T,1Hz). The normalization of the relative change of impedance was done with stable impedance signals (Z_PBS,1Hz) in PBS (pH 7.2). Impedance was normalized to make impedance sensors independent of electrode size or active surface area throughout each biofilm experiment in PBS (pH 7.2).

2.4. Dental plaque biofilm growth on glass and resin substrate.

To quantify the dental plaque-derived multispecies biofilm volume and the corresponding chemical microenvironment change, we grew dental plaque derived-multispecies biofilm on the relevant substrates beside the impedance and pH sensors (see Figure S1). The substrates were placed in the 30 mm Petri dish, as shown in Figure S1A. Substrates preparation protocol can be found in the supporting information.

Impedance and pH sensors (Figure S1B &C) with pH control sensors were all attached to the substrate, as shown in Figure S1D, inside the incubator box. One pH and one control electrode were placed in the bulk solution to monitor pH change in bulk solution with biofilm growth. The BMM growth media was passed through the dish to record baseline signals of all the sensors. Once the baseline signals were stabilized after ~12 h, the flow of the solution was stopped, the freshly grown dental plaque-derived multispecies bacterial in BMM having OD₆₀₀ ~ 0.1 (stock sample OD₆₀₀ ~1.0, sample collection protocol approved IBC#6120)-was injected and inoculated for 2 h. The inoculation sample was prepared in the presence of sterile 30 mM glucose. The flow of the BMM growth media was resumed after inoculation. The bacteria within the dish were fed with 30 mM glucose every 8 h for the entire experimental period. The biofilms were grown inside the incubator box at 37°C in the presence of 5% CO₂ gas flow. The flow rate of all solutions was maintained at 3.6 mL/h for the inlet and 3.3 mL/h for the outlet. The biofilm was washed with BMM every 24 h after 2 days of inoculation. The impedance signals were recorded with biofilm formation using impedance signals at every 20 minutes throughout the experiments. The potentials change of pH and control signals were recorded continuously every 30 seconds. All the studies reported here were done in triplicate. Resin and BAG-composite fabrication were prepared through sol-gel process as previously described [32,33].

2.5. Biofilm volume determination by CLSM.

We chose to scan 20 locations (randomly selected), each area being 0.25 mm² on the exposed surface area of the substrates. Protocol details can be found in the supporting information.

3. RESULTS AND DISCUSSION

3.1. Characterization of PEDOT electrodes as impedance sensors.

We deposited PEDOT: PSS on the Pt wire electrode for the impedance sensor to monitor dental plaque biofilm in real-time because of its favorable electrochemical properties, biocompatibility, and stability. The PEDOT: PSS was deposited through cyclic voltammetry, as shown in Figure S2(A and B), and the quality of the deposition was checked with 1 mM ferrocene methanol, as shown in Figure S2(C and D). Further, we compared the electrical properties of the impedance sensor (PEDOT electrode) and bare Pt. The charge transfer resistance was calculated with a fitted equivalent circuit (Table S1) from the impedance experiments (Figure 1A and B), suggesting that there is an approximately four-fold (3.88 ± 0.74 (n=3)) decrease in charge transfer resistance for the impedance sensor compared with that for the bare Pt. Hence, the PEDOT electrode showed improved electrical conductivity in comparison to that of the bare Pt electrode.

Figure 1.

Electrochemical impedance spectroscopy measurement in frequency range of 100 kHz to 0.1 Hz at 0.1 V (vs Ag/AgCl), using 5 mV sinusoidal excitation wave. (A) Nyquist plot and (B) Bode plot of Pt wire and impedance sensors in 1mM ferrocenemethanol (FcMeOH). Impedance sensor characterization in frequency range of 100 kHz to 1 Hz at 0.1 V (vs Ag/AgCl), using 5 mV sinusoidal excitation wave. (C) Nyquist plot and (D) Bode plot of impedance sensors in PBS and BMM.

Moreover, it has been reported that that PSS doped PEDOT is porous structure [34]. The effect of PEDOT: PSS coating on Pt wire is shown in Figure 1(A &B). In PEDOT: PSS coated Pt wire, there are small semicircle (Figure1A) and non-zero phase angle at high frequency (100kHz) (Figure 1B) as compared to Pt wire. The semicircle in Nyquist plot and phase hump in Bode plot at high frequency has been reported for mesoporous and micro-mesoporous structure. These results are attributed due to comparability of average mesoporous surface curvature with the dynamic diffuse screening layer thickness and that affects the electrical double layer (EDL) relaxation time constant [35]. The impedance signal suggests that PEDOT: PSS coating is micro-mesoporous or mesoporous in nature.

To further investigate the reproducibility and sensitivity of impedance sensors, we characterized the PEDOT electrodes or impedance sensors in PBS and BMM. As shown representatively in Figure 1(C &D), the equivalent circuit [36] fitted data (Table S2) showed the reproducible characteristic double-layer capacitance of the sensor to be 62 ± 6 μF (n=9) for PBS and 45 ± 3 μF (n=3) for BMM. The sensitivity of impedance sensor is characterized in terms of improved electroactive surface area (ECSA) and electrochemical signal (Figure S2C and D). The electroactive surface area (ECSA) of impedance sensor was increased ~ three-folds of magnitude as compared to the geometric surface area. The large ECSA of impedance sensor will facilitates protein adsorption and bacteria attachment. The improvement of ECSA causes three folds of increase in sensitivity of oxidation current of ferrocene methanol as compared to bare Pt wire (Figure S2C and D). We concluded that the PEDOT-coated Pt wire can act as a sensitive and reproducible impedance sensor that is capable of performing in protein-rich complex bacterial growth media such as BMM.

The developed impedance sensor (PEDOT: PSS coated Pt wire) is one time usable because of electrode surface modification after adsorption of proteins, bacteria, and EPS matrix. However, Pt wire can be reused after cleaning of PEDOT: PSS layer and redeposit new PEDOT: PSS layer. The impedance sensors can be used within 7 days of electrode fabrication. The sensors are stable in biofilm growth conditions for ~ 12 days.

3.2. Optimization of frequency of impedance signal to quantify the biofilm volume in real-time.

After fabricating and optimizing the PEDOT: PSS electrode so that it could function as an impedance sensor, we optimized the frequency of the impedance signal to quantify the biofilm volume in real-time. In this study, we aimed to develop an impedance-based analytical method to quantify the entire biofilm, including the polysaccharide matrix, rather than a single bacteria suspended in solution. Therefore, we collected the impedance data (100kHz to 1Hz) approximately every 12 h for the optimization of operational single frequency of impedance sensors as shown in Figure 2(A–C). The impedance signals at 1Hz frequency are highly sensitive in terms on S/N ratio with biofilm growth time as compared to other frequencies signals (Figure 2D). At 1Hz frequency, the impedance signal is more sensitive to the change of double layer capacitance as phase angle is high (~80°) (Figure 2C). The double layer capacitance is highly sensitive to the bacterial attachments and biofilm growth on electrode surface (as showed representative in Figure 2E). Therefore, we choose to collect impedance data at 1Hz in all experiments. Most of reported work in the literature are based on the detection of bacteria based on the change of capacitance and resistance of electrodes with bacteria attachments at high frequency of impedance signals [27,37–39].

Figure 2.

The representative data for optimization of frequency of electrochemical impedance spectroscopy (EIS) spectra to monitor biofilm volume in real-time through different stages of biofilm growth inside the box incubator on resin substrate. (A) Nyquist plot (B) Bode plot, (C) phase angle Bode plot. EIS was performed at 0.1 V (vs Pt wire/PEDOT) in frequency range from 100 kHz to 1 Hz with amplitude 5 mV. (D) % change of impedance signals with biofilm growth time at different frequencies. (E) Schematic setup of impedance sensors inside the biofilm.

3.3. Impedance sensor with dental plaque-derived multispecies biofilm.

The biofilm experiment was set up inside the incubator box as shown in Figure S1. The total impedance signals (Z_Total) of impedance sensors was monitored every ~20 min. We have used PBS in initial stage of experiments to check stability of the impedance sensors. The impedance signal stabilized in the presence of PBS within 12 h as shown in Figure 3A. PBS was then switched to BMM as growth media for biofilm formation. The signal noise in Figure 3A is due to change of PBS to BMM. The impedance sensors took an additional ~12 h to stabilize the impedance signal (Figure 3A). The multispecies derived dental plaque bacteria (OD₆₀₀ ~ 0.1) were injected into the dish after stabilization of baseline signals in BMM, and the time was reset to zero, as shown in Figure 3B.

Figure 3.

Biofilm volume and impedance relationship at 1 Hz frequency for glass substrate. (A) Representative data of impedance measurements over time with different solutions of PBS, BMM, and BMM and bacteria biofilm. (B) Representative data of normalized Z values over time. (C) Calibration curve established with biofilm volume obtained from the confocal data at three different percentage changes in impedance with a sensitivity of 0.60 ± 0.03 biofilm volume / % change in impedance and an intercept of −1.3 ± 0.073 (n=4 for each percentage). (D) The representative data of converted biofilm volume in real-time corresponding to the impedance values.

To obtain the impedance signal from bacteria attachments and biofilm growth in BMM growth media, we have used impedance signal in BMM as baseline signals. The normalization of impedance signals has been calculated using baseline signals in BMM and stable signals in PBS. A representative sample of the normalized percentage change in impedance over time for the multispecies biofilm is shown in Figure 3B. We observed fluctuations in the impedance data at 1.5, 2.5, 3.5- and 5-days’ time stamps, which might be due the enhance metabolic activity during the bacteria glucose feeding cycles.

To quantify the biofilm volume, we developed a calibration curve following IUPAC guidelines correlating the impedance change vs biofilm volume on a standard glass substrate [40]. As shown in Figure 3C, changes in impedance of 2.5, 35, 50, and 65% were chosen to represent four different stages of biofilm growth: initial bacteria incubation, early attachment, growth phase, and mature biofilm. For each percentage, the experiment was stopped, then the biofilm was taken out. Afterward, its volume measured by CLSM fluorescent imaging at 20 different locations of the 3 mm diameter biofilm to avoid any inherent bias toward collecting the confocal data (70% of total geometric area was assessed). Afterward, the biofilm volume was normalized by total geometric area (5 mm²) and linked to the corresponding percentage change in impedance. The working calibration curve obtained was y = (0.60 ± 0.03) x − (1.3 ± 0.073) (Equation 1) (n=4 for each percentage) with R² value 0.99. The equation was then used to convert the impedance signal to biofilm volume in real-time, as shown in Figure 3D.

To demonstrate the universality of the method in estimating the volume of multispecies biofilm formed on additional substrates (Figure 3C), we adapted it to monitor multispecies biofilm growth on resin or dental filling material. As in the earlier experiment, a calibration curve was generated for a multispecies biofilm formed from dental plaque in order to correlate the volumes and percentage change in impedance across a 3 mm diameter exposed resin region (Figure 4A). The equation for resin is reported to be y = (0.50 ± 0.03) x + (0.93 ± 0.53) (Equation 2) (n=4 for each percentage). There was no statistically significant difference between the two linear regressions for standard glass and resin substrates, using linear regression model for interaction (R studio) at the 95% confidence interval (p-value = 0.32). As a result, we determined that the impedance sensor-based technique is substrate independent. Although there were no substantial variations in the operating equation, there were variances in the growth kinetics of the bacterium biofilm on two different substrates.

Figure 4.

Validation of impedance-based method with different substrates and kinetic growth of biofilm. (A) Calibration curve for the resin substrate established with biofilm volume obtained from the confocal data at three different percentage changes in impedance with a sensitivity of 0.50 ± 0.03 biofilm volume / % change in impedance and an intercept of 0.93 ± 0.53. (n=4) (B) Representative data of converted biofilm volume in real-time corresponding to the impedance values, using a resin calibration curve for the resin substrate. (C) Representative data of converted biofilm volume in real-time corresponding to the impedance values, using a glass calibration curve for BAG-composite. (D) Comparison of kinetic growth with the same percentage change in impedance between three different substrates.

3.4. Effect of Ca²⁺-releasing resin-BAG composites on biofilm volume.

We aimed to further validate the working calibration curve of biofilm volume vs percentage change in impedance by using a novel Ca²⁺-releasing resin-BAG dental composite as the testing substrate. We chose an arbitrary point on the working calibration curve, such as a 60% change in impedance, and then calculated the anticipated biofilm volume on the BAG-composite substrate by using the working equation (Equation 1) (shown in Figure 4C). The calculated volume was 35 μm³/μm². We validated the biofilm volume experimentally by using the previously known confocal fluorescence image analysis methods. The biofilm volume was determined to be 37 ± 2.3 μm³/μm² (n=4). The percentage error between experimental data (37 ± 2.3 μm³/μm²) and estimated data (35 μm³/μm²) was 6%. The impedance-based method was shown to be credible for future volume prediction on various substrates with a 6% error for a biological system. The biofilms that developed on glass, resin, and BAG-composite required 3.5, 4.5 and 6 days, respectively, to achieve a 50% change in impedance. In comparison to the glass substrate, the resin and BAG-composite substrate exhibited a delay in development at the 50% change in impedance, as illustrated in Figure 4D. In addition, the impedance data demonstrated a considerable delay in the initial adhesion of bacteria and the growth of biofilms on BAG-composite substrates. The difference of biofilm formation kinetics has been affected by surface composition, roughness and porosity of substrates. The glass substrate is smoother than resin and BAG-composite. The difference in the initial attachment of bacteria and biofilm formation on glass and resin substrates suggests that the surface roughness and porosity delayed the bacteria attachment and initial biofilm formation as shown in Figure 4D. Even though resin and BAG-composite substrates have been polished to make same roughness and porose surface, initial bacteria attachment and biofilm growth kinetics are different due to different surface composition (Figure 4D). We expect that Ca²⁺ release from BAG-composite substrates cause delay of initial bacteria adsorption and biofilm formation [32].

We also determined bacterial distribution within the biofilm at various positions on the substrate by using a live/dead assay, as shown in Figure S3. In comparison to the BAG-composite substrate, the bacteria were more equally distributed across the glass or resin substrates. At some locations, the BAG-composite biofilm matrix was more dense, globular, and heterogeneous (Figure S3), which may have an effect on the diffusion of ions, including nutrients and metabolites produced by bacteria. With the newly developed impedance-based method, we are able to monitor the biofilm volume growth in real-time with various substrates including glass, dental resins, and BAG-composite substrates. Most studies were reported on the attachment of bacteria and biofilm growth on a conductive/electrode material instead of bio non-conductive materials. Biofilm volume quantification and monitoring continuously in real-time were not investigated for various substrates in these studies [26,27,37,41].

3.5. Correlation of pH and biofilm volume on different substrates.

We used a flexible micro pH sensor at the interface between the substrate and the multispecies biofilm in each experiment to monitor the bacteria’s organic acid production and to investigate the relationship between the rate of biofilm formation and the resulting change in local pH between the biofilm and the substrate. Three distinct substrates, including glass, resin, and BAG-composite, were used in this study to monitor the change in local pH over time while biofilms grew on them (as shown representative in Figure S4). We observed that when the biofilm volume reached 18 μm³/μm² on any substrate (Figure S4), the local interfacial pH decreased from pH 7 to 6.4 ± 0.20. This observation indicates that when the volume of the biofilm reaches 18 μm³/μm², it can produce a wide range of adequate organic acids, including lactic acid, through carbohydrate metabolism to titrate the buffer and lower the pH with 0.58 ± 0.20 (n=8) units for these substrates (Figure S4A–C) [42]. It is worth mentioning, however, that the rate at which the volume of the biofilm changes and the resulting pH changes differs by substrate. For example, biofilms grown on glass or resin required 3.0–3.5 days to reach a volume of 18 μm^3/μm², whereas biofilms grown on a BAG-composite substrate required nearly 5 days to reach the same volume, which modified the local pH. This also shows that substrates have an effect on the rate of biofilm formation and may have played a key role in these observations.

3.6. Biofilm and equivalent circuit model.

As described previously, we established impedance spectroscopy as an analytical technique for the real-time quantification of living biofilms on various substrates. We further investigated the relationship between the impedance signal and various stages of biofilm growth by using an impedance equivalent circuit model. After converting the impedance value to the biofilm volume, we examined three distinct growth stages, based on time and biofilm volume, in order to gather their complete impedance spectra, as shown in Figure 5A. Later, the impedance spectra were fitted by using CHI workstation (model no. 760E, CH Instruments) for equivalent circuit modeling. Several equivalent circuit modeling for biofilm formation have been reported in the literature [37–39,43]. The insulating nature of bacteria and biofilm matrix have been modeled with biofilm resistance and biofilm capacitance elements in parallel configurations. Bobacka et. al. has proposed equivalent circuit for PEDOT:PSS materials without bacteria [36]. Later, Gkoupidenis and his group have modified previous model and proposed new equivalent circuit for bacteria attachment on PEDOT:PSS coated metal electrode [21]. In this study, we have combined these two proposed circuits for PEDOT: PSS and given new equivalent circuit (Figure 5B). The five equivalent circuit parameters define the biofilm formation. R_s is solution resistance, constant phase element (CPE) attributed from capacitance caused by proteins, bacteria attachments, and biofilm formation, R_b is biofilm resistance caused by insulation bacteria and biofilm formation, C_P denotes the sensor’s inherent capacitance, and Z_w is linear ion diffusion through biofilm (Figure 5B). Since, we have used two identical electrodes as impedance sensors for biofilm monitoring, single R_b, CPE and C_P have been used for simplification in data fitting. Percentage change of biofilm resistance (R_b) has been used to compare early, middle, and mature state of biofilm formation. The increase of R_b in early stage is 57 (±30) % with respect to BMM. The R_b has been increased 168 (±50) % and 536 (±118) % for middle and mature state of biofilm respectively. Since, R_b is in parallel configuration with other equivalent circuit elements, even though 536% increase in R_b does not contributes significantly to Z_total. The R_b at early, middle, and mature state of biofilm are significantly different at 95% confidence intervals (p-value<0.05). The increase in biofilm resistance was a result of the volume and density of the biofilm matrix increasing with time. The combination of resistor parameters in series with each bacterium and the attachment of the extracellular polymeric substance (EPS) matrix can be used to visualize the increase in biofilm resistance [41] As shown in Figure 3C, the linear relationship between biofilm resistance (or total impedance) and biofilm volume was established. Biofilm capacitance decreased with biofilm growth time from early to mature stages (Table S3 [37] (Figure 5A). The decrease in biofilm capacitance is expected due to an increase in biofilm thickness (biofilm volume) [44]. The capacitance of the biofilm in series with each bacterium and the EPS matrix layer results in a decrease in the equivalent capacitance. During the development phase, bacteria adhere gradually, and the EPS matrix accumulates, affecting the biofilm’s resistance and capacitance (Figure 5C). The increasing trend of R_b and the decreasing trend of C_b suggest that mature biofilms form gradually. Thus, the equivalent circuit may be used to predict the various stages of biofilm growth, including initial bacterial adhesion, biofilm creation, and finally, biofilm maturation.

Figure 5.

The equivalent circuit and biofilm biological complex system. (A) Representative data of Nyquist plot of different stages of biofilm growth on glass substrate. (B) Proposed equivalent circuit model of the biofilm-impedance sensor system according to the collected data. (C) Schematic diagram of the complex biological process on top of the impedance sensor.

4. CONCLUSION

In this study, real-time impedance and biofilm volume measurements revealed slower kinetics of biofilm formation and development for a BAG-composite than for glass and resin substrates. This observation was enabled by our newly designed and well-established noninvasive impedance-based technique which can monitor real-time biofilm growth on various substrates. In addition, we used pH sensor to support the association between the pH microenvironment and biofilm formation. When biofilm growth reaches 18 μm³/μm² biofilm volume on these substrates, sufficient organic acid is produced to titrate the buffer within the biofilm, which causes the change of pH at the biofilm and substrates interface. The standardized impedance-based method for real-time monitoring of biofilm formation from impedance-based method and local pH from the potentiometric sensor would be beneficial for researchers and clinicians interested in determining how different materials’ surfaces affect biofilm growth.

Highlights.

• A noninvasive impedance approach for real-time monitoring of dental plaque-derived multispecies biofilm growth has been developed.

• This impedance-based technique could be useful for real-time monitoring of biofilm growth on a variety of substrates.

• The kinetics of biofilm growth differ between glass, resin, and BAG-Ca²⁺ substrates.

• pH wire sensors were used to determine the local pH change at the biofilm-substrate interface.

• This method could be beneficial for investigating bacteria-mediated corrosion or antibacterial properties of novel biomaterials.

Biographies

Dr. Dipankar Koley is an associate professor in the department of chemistry at Oregon State University. Dr. Koley obtained his Ph.D. in chemistry from the University of Texas at Austin in 2011 under the guidance of Dr. Allen J Bard. Later, Dr. Koley moved to the University of Michigan as a post-doctoral fellow to work under Dr. Mark E Meyerhoff. His research interest lies at the intersection of electrochemistry, biology, and bioengineering. In OSU, Dr. Koley and his team is working on developing new electrochemical techniques to understand the microbial metabolic exchange in biofilm at high spatial and temporal resolution.

Jack Ferracane is a professor and head in restorative dentistry in Oregon Health and Science University. His research involves in developing new dental composites to address the secondary caries, a major problem in the field of dentistry.

Sriram Kumar is a post-doctoral fellow working under the supervision of Dr. Dipankar Koley, Department of Chemistry.

Anh Tuan Nguyen is a Ph.D. student working under the supervision of Dr. Dipankar Koley, Department of Chemistry, Oregon State University, Oregon, USA. His research is mainly focused on the study of bacterial metabolism using electroanalytical techniques.

Subir Goswami is a visiting scientist in Department of chemistry, Oregon State University.