Genetic engineering biofilms in situ using ultrasound-mediated DNA delivery

Abstract

The ability to directly modify native and established biofilms has enormous potential in understanding microbial ecology and application of biofilm in 'real-world' systems. However, efficient genetic transformation of established biofilms at any scale remains challenging. In this study, we applied an ultrasound-mediated DNA delivery (UDD) technique to introduce plasmid to established non-competent biofilms in situ. Two different plasmids containing genes coding for superfolder green fluorescent protein (sfGFP) and the flavin synthesis pathway were introduced into established bacterial biofilms in microfluidic flow (transformation efficiency of 3.9 ± 0.3 × 10^-7 cells in biofilm) and microbial fuel cells (MFCs), respectively, both employing UDD. Gene expression and functional effects of genetically modified bacterial biofilms were observed, where some cells in UDD-treated Pseudomonas putida UWC1 biofilms expressed sfGFP in flow cells and UDD-treated Shewanella oneidensis MR-1 biofilms generated significantly (P < 0.05) greater (61%) bioelectricity production (21.9 ± 1.2 µA cm^-2 ) in MFC than a wild-type control group (~ 13.6 ± 1.6 µA cm^-2 ). The effects of UDD were amplified in subsequent growth under selection pressure due to antibiotic resistance and metabolism enhancement. UDD-induced gene transfer on biofilms grown in both microbial flow cells and MFC systems was successfully demonstrated, with working volumes of 0.16 cm³ and 300 cm³ , respectively, demonstrating a significant scale-up in operating volume. This is the first study to report on a potentially scalable direct genetic engineering method for established non-competent biofilms, which can be exploited in enhancing their capability towards environmental, industrial and medical applications.

Fig. 1

Schematic diagram of ultrasound‐based DNA delivery (UDD) into bacterial cells of mature biofilms established in (A) microfluidic flow cells and (B) microbial fuel cell (MFC). Ultrasound treatments were applied in a commercially available 42 kHz ultrasound cleaning bath. Diagram is not drawn to scale.

Fig. 2

Time course of ultrasound‐mediated gene transfer into a biofilm. Green fluorescence signal (ex. 485nm, em. 510nm) and bright‐field imaging of biofilm samples after 120 h with (A) the addition of both plasmid and ultrasound treatment (+P/+U), (B) only the addition of plasmid (+P/−U), (C) only ultrasound treatment (−P/+U) and (D) no plasmid and no ultrasound treatment (−P/−U). Green fluorescence signal and bright‐field imaging for biofilm samples with both addition of plasmid and ultrasound treatment after (E) 5 h, (F) 24 h, (G) 48 h and (H) 120 h.

Fig. 3

(A) Electric current density I versus elapsed time, (B) polarization curve (current density vs. potential) and (C) power density curve of MFC reactors with S. oneidensis MR‐1 WT (blue), MR‐1/YYDT‐C5 mutant (orange) and MR‐1 Δbfe strains (grey) with 20 mM sodium lactate as sole carbon source. Measurements were conducted via Linear Sweep Voltammetry, as described above. Error bars represent the standard deviation of triplicate measurements. (D) Measured optical density at 600nm (OD₆₀₀) of anodic culture of MFC reactors utilizing S. oneidensis MR‐1 WT, MR‐1/YYDT‐C5 mutant and MR‐1 Δbfe strains with 20 mM sodium lactate as sole carbon source. Measurement was done using 1 cm cuvette (1 ml sample size). (E) Biofilm quantification using crystal violet assay: optical density at 595 nm (OD₅₉₅) of cell‐bound crystal violet solution from anodic biofilm cells of the MFC reactors. F. The amount of lactate consumed by each reactor. Produced metabolites were mainly acetate, with succinate and pyruvate in trace amounts (data not shown). Measurements in Figure (D), (E) and (F) were done at the end of MFC experiment (day 13). Error bars represent standard deviation of triplicate measurements. P values on top of the bars denote differences between sample pairs based on nested mixed‐factor ANOVA test followed by Tukey’s HSD post hoc test. P values showing statistically significant (P < 0.05) differences are presented in bold.

Fig. 4

(A) Electric current densityIof double‐compartment MFC reactors running at 1kΩ load with 20 mM initial concentration of sodium lactate; (B) extracellular flavins concentration of MFC reactors after 14 days of operation. Four different type of reactors: MR‐1/YYDT‐C5 strain (MR‐1/YYDT‐C5_US, orange), MR‐1 WT with addition of plasmid and ultrasound treatment (WT_P_US, blue), MR‐1 WT with only ultrasound treatment (WT_US, yellow) and WT with only addition of plasmid (WT_P, grey). Ultrasound was performed for 30s on day 6 (black arrow) for appropriate MFC set‐ups. On day 9, kanamycin (10 µg ml^‐1) and 10 mM of additional lactate were added into each reactor (light blue arrow). On day 13, additional kanamycin was added to reach final concentration of 50 µg ml^‐1 (green arrow). Shaded regions represent standard deviations of triplicate measurements. P values on top of the bars were calculated for the last day of measurement and denote differences between sample pairs based on nested mixed‐factor ANOVA test followed by Tukey’s HSD post hoc test. P values showing statistically significant (P < 0.05) differences are presented in bold.