Real-time monitoring of bacterial biofilms metabolic activity by a redox-reactive nanosensors array

Abstract

Background: Bacterial biofilms are communities of surface-associated microorganisms living in cellular clusters or micro-colonies, encapsulated in a complex matrix composed of an extracellular polymeric substance, separated by open water channels that act as a circulatory system that enable better diffusion of nutrients and easier removal of metabolic waste products. The monitoring of biofilms can provide important information on fundamental biofilm-related processes. That information can shed light on the bacterial processes and enable scientists to find ways of preventing future bacterial infections. Various approaches in use for biofilm analysis are based on microscopic, spectrochemical, electrochemical, and piezoelectrical methods. All these methods provide significant progress in understanding the bio-process related to biofilm formation and eradication, nevertheless, the development of novel approaches for the real-time monitoring of biochemical, in particular metabolic activity, of bacterial species during the formation, life and eradication of biofilms is of great potential importance.

Results: Here, detection and monitoring of the metabolic activity of bacterial biofilms in high-ionic-strength solutions were enabled as a result of novel surface modification by an active redox system, composed of 9,10-dihydroxyanthracene/9,10-anthraquinone, on the oxide layer of the SiNW, yielding a chemically-gated FET array. With the use of enzymatic reactions of oxidases, metabolites can be converted to H₂O₂ and monitored by the nanosensors. Here, the successful detection of glucose metabolites in high-ionic-strength solutions, such as bacterial media, without pre-processing of small volume samples under different conditions and treatments, has been demonstrated. The biofilms were treated with antibiotics differing in their mechanisms of action and were compared to untreated biofilms. Further examination of biofilms under antibiotic treatment with SiNW-FET devices could shed light on the bioprocess that occurs within the biofilm. Moreover, finding proper treatment that eliminates the biofilm could be examined by the novel nanosensor as a monitoring tool.

Conclusions: To summarize, the combination of redox-reactive SiNW-FET devices with micro-fluidic techniques enables the performance of rapid, automated, and real-time metabolite detection with the use of minimal sample size, noninvasively and label-free. This novel platform can be used as an extremely sensitive tool for detection and establishing medical solutions for bacterial-biofilm eradication and for finding a proper treatment to eliminate biofilm contaminations. Moreover, the sensing system can be used as a research tool for further understanding of the metabolic processes that occur within the bacterial biofilm population.

Keywords: Bacteria; Biofilms; Field effect transistors; Metabolism; Nanosensors; Silicon nanowires.

Fig. 1

Microfluidic redox-reactive nanoFET biosensor for extracellular bacterial metabolic analysis. a Silicon wafer chip, with 600 nm thermal oxide layer, which contains 200 potential redox-reactive SiNW FET devices, sharing a common gate. The nanoFETs covered with PDMS microfluidic channel connected via tubing to an Eppendorf tube with a small bacterial media sample mixed with oxidase enzyme. The forming bacterial biofilms of B. subtilis are shown in the left panel. Inset: scanning electron microscope image of single redox-reactive nanoFET consisting of 20 nm p-type SiNW connected to the source and drain electrodes. The nanoFETs chip wire-bonded to the PCB holder, which is connected to the electrical recording system (Additional file 1: Figure S13). b Operation mechanism of the redox-reactive nanoFET biosensor. The redox-reactive nanoFET biosensor reversely reduced or oxidized in the present DEHA or H₂O₂, respectively. When the redox reactive device is oxidized, the conductivity of the device increased, and AQ moieties are formed on the nanoFET surface (right panel). On the other hand, when the redox reactive device is reduced, the conductivity of the device decreased and DHA moieties are formed on the nanoFET surface (left panel)

Fig. 2

Comparison between optical and redox-reactive nanowire FET measurements of bacteria’s metabolic activity. a Transmittance versus time, and b Glucose signal versus time, as measured by spectrophotometer (at 600 nm) and by redox-reactive SiNW, respectively, during the growth of E. coli bacteria. in minimal broth medium supplemented with glucose as the only carbon source, at pH 7.3. inset of b Demonstration of the calculation method used to extract plot b. Before each injection of a new sample (800 µL sample, rate = 100 µL/second), the devices were switched off (Vsd = 0 V). Each measurement lasted for 180 s, during which the source-drain voltage (Vsd) was 0.3 V and the gate voltage (Vg) was 0 V. The glucose signal-versus-time curve was constructed from the current values at 180 s from the beginning of each measurement, right before the devices were switched off (Vsd = 0). The sensing conditions were chosen based on previous studies [13]. The signal was calculated by subtracting the value of the current after the addition of oxidase from the value of the current before its addition. The black dashed line marks the baseline obtained by the addition of reductant (1 vol% DEHA in medium). The error in the current values (Y-axis in b) is estimated to be ± 0.100 nA, the error in the transmittance values (Y-axis in a) is calculated as a function of the absorption measurements, as described in Additional file 1: Section 15

Fig. 3

Monitoring of the metabolic activity of bacterial biofilm. The metabolic activity was monitored with the use of a redox-reactive SiNW-FET device and the time-dependence of the glucose signal is shown. First, the biofilm was grown in MSgg medium (glycerol and glutamate) and incubated at 30 °C for 40 h. Then, the MSgg medium was replaced by the glucose-based medium. The bacterial biofilm consumes all the glucose in 11 h. The minimal-broth medium contains 5.5 mM glucose at the beginning of the measurements. Inset: bacterial-growth rate curve presented as the logarithm of glucose signal versus time. The lag phase and log phase of bacterial glucose metabolism are monitored with the use of a redox-reactive SiNW FET

Fig. 4

Bacterial biofilm was monitored over two cycles of glucose consumption. First, the bacterial biofilm is incubated with a minimal-broth medium containing 1.8 mM glucose, and then the medium was renewed twice. Each consumption cycle lasted approximately 150 min

Fig. 5

Bacterial biofilm was monitored over three cycles of glucose consumption; untreated biofilm (blue), short-term exposure to tetracycline (red), and long-term exposure (black). The long-term treatment decelerates the metabolic cycle and, in the end, ~ 20% of the glucose remains. At the beginning of each cycle, the glucose medium contains 1.8 mM glucose. Inset: trend lines of the three glucose-consumption cycles

Fig. 6

Bacterial biofilm was monitored over three cycles of glucose consumption; untreated biofilm (blue), short-term exposure to ampicillin (red), and long-term exposure (black). The long-term treatment slows the metabolic cycle and, in the end, the glucose signal shows ~ 40% remaining of the original amount. At the beginning of each cycle, the glucose medium contains 1.8 mM glucose. Inset: trend lines of the three glucose-consumption cycles

Fig. 7

Bacterial biofilm was monitored over two cycles of glucose consumption; untreated biofilm (blue) and bacterial biofilm exposed to UV irradiation of 356 and 405 nm with intensities of 9.3 and 21.6 mW/cm², respectively, for 10 min. Inset: a trend line of the two glucose-consumption cycles