All the bacterial strains and plasmids involved in this research are listed in Table 1. Escherichia coli (E. coli) TOP10 was used as a host strain for cloning and biosensing. Genetically-engineered bacteria were grown at 37^oC in Luria-Bertani (LB) broth containing 1% tryptone (Oxoid, Basingstoke, United Kingdom), 0.5% yeast extract (Oxoid, Basingstoke, United Kingdom), and 1% sodium chloride supplemented with 50 μg/mL ampicillin. DNA primers and fragments were synthesized by Sangon Biotech (Shanghai, China). Restriction endonucleases, reagents, and buffers used in molecular cloning were purchased from Sangon Biotech. All other chemicals were of analytical grade and were obtained from Sigma-Aldrich (St Louis, MO, United States). Stock solutions of cadmium chloride, mercuric chloride, lead nitrate, and zinc sulfate were freshly prepared using ultrapure water and filtered through a 0.22 μM filter. All recombinant plasmids were identified by colony PCR, restriction enzyme digestion, and DNA sequencing.

The strategy used for the assembly of biosensing constructs is shown in Figure 1A. Recombinant plasmids were assembled by a combination of PCR and subcloning techniques. An 887 bp fragment containing the promoterless egfp gene and rrnB terminator was PCR amplified from the pT-GFP vector and subcloned into the pPmer vector using the NdeI and HindIII restriction sites to generate pPmer-G. Similarly, an 878 bp fragment containing the promoterless mcherry gene and rrnB terminator was PCR amplified from the pT-RFP vector and subcloned into the pPcad vector using the NdeI and XhoI restriction sites to produce pPcad-R. A DNA fragment containing a mCherry open reading frame (ORF) was amplified from pT-RFP and fused with the Hg(II) sensory element merR-Pmer by an overlapping extension PCR as described previously (Guo et al., 2021a) to generate pR-Pmer-G.

Similarly, a DNA fragment containing an eGFP ORF was amplified from pT-GFP and fused with the Cd(II) sensory element cadR-Pcad by an overlapping extension PCR to generate pG-Pcad-R. The cadR-Pcad-mcherry cassette was PCR amplified from pPcad-R and reversely inserted into the BglII and BamHI sites of pET-21a to generate pR-Pcad. The merR-Pmer-egfp cassette was PCR amplified from pPmer-G and inserted into the BamHI and HindIII sites of pR-Pcad to generate pR-Pcad-Pmer-G, which was designed as a dual-sensing construct to detect bioavailable Cd(II) and Hg(II) simultaneously. All the biosensing constructs were used for transformation into E. coli TOP10 competent cells and bacterial transformants harboring the biosensing constructs were selected on LB agar plates containing 50 μg/mL ampicillin.

A potential mechanism for simultaneous differential detection of bioavailable Cd(II) and Hg(II) based on a dual-sensing system is shown in Figure 1B. When only Cd(II) exists in bacterial cytoplasm, selective binding with dimeric CadR is mediated by the cooperation of thiolate-rich and histidine-rich sites, leading to the conformational change of dimeric CadR (Liu et al., 2019). The resultant activated dimeric CadR is critical for enhancing the transcriptional activity of downstream red fluorescent reporter. Although some studies have shown dimeric CadR could also be activated by intracellular Hg(II) (Tao et al., 2013; Hui et al., 2021c), dimeric MerR will undoubtedly become the dominant transcription activator to trigger the transcription of green fluorescence protein upon exposure to only Hg(II). Both red and green fluorescence will be emitted simultaneously when Cd(II) and Hg(II) coexists.

Single colonies of various single-sensing whole-cell biosensors were picked from the agar plates and used to inoculate 3 mL of LB medium and cultured at 37^oC for 12 h. Overnight cultures were inoculated into fresh LB medium at 1% inoculation amount. To investigate the bidirectional transcription levels of TOP10/pR-Pmer-G and TOP10/pG-Pcad-R upon exposure to their cognate metal ions, TOP10/pR-Pmer-G in lag phase was directly exposed to 0-10 μM Hg(II), and TOP10/pG-Pcad-R was exposed to 0–100 μM Cd(II). To investigate the influence of the cloning direction on the response of the Cd(II)-sensing module, TOP10/pPcad-R, and TOP10/pR-Pcad were exposed to 0–100 μM Cd(II). After culturing at 37^oC for 8 h, bacterial cell density and fluorescent signals were measured.

Overnight cultures of single-sensing TOP10/pPmer-G, single-sensing TOP10/pR-Pcad, and dual-sensing TOP10/pR-Pcad-Pmer-G were inoculated into fresh LB medium at 1% inoculation amount. Stock solutions of Zn(II), Pb(II), and Cd(II) were added at a final concentration of 5, 25, 125, or 250 μM, and stock solution of Hg(II) was added at a final concentration of 5 or 25 μM. After culturing at 37^oC for 8 h, bacterial cell density and fluorescent signals were measured.

Overnight cultures of single-sensing TOP10/pPmer-G, single-sensing TOP10/pR-Pcad, and dual-sensing TOP10/pR-Pcad-Pmer-G were inoculated into fresh LB medium at 1% inoculation amount. A double dilution method described previously (Hui et al., 2021c) was used to obtain 40, 20, 10, 5, 2.5, 1.25, 0.625, 0.3125, 0.156, 0.078, 0.039, 0.0195, 0.0098, 0.0049, 0.0024, and 0 μM Hg(II) exposure groups. The resultant cultures were grown at 37^oC for 8 h before measuring bacterial cell density and fluorescent signals.

Overnight cultures of single-sensing TOP10/pR-Pcad and dual-sensing TOP10/pR-Pcad-Pmer-G were inoculated into fresh LB medium at a 1% inoculation amount. A double dilution method was used to obtain 800, 400, 200, 100, 50, 25, 12.5, 6.25, 3.125, 1.56, 0.78, 0.39, 0.195, 0.098, 0.049, 0.024, 0.012, and 0 μM Cd(II) exposure groups. The resultant cultures were grown at 37^oC for 8 h before assessing the fluorescent signals and bacterial cell density.

To study the influence of Cd(II) on the response of the dual-sensing TOP10/pR-Pcad-Pmer-G toward Hg(II), the mixtures of 2.5 μM Hg(II) with 0–400 μM Cd(II) were added into the cultures of TOP10/pR-Pcad-Pmer-G during the lag phase. Double-color fluorescent signals and bacterial density were determined after an 8-h induction at 37^oC.

To evaluate the influence of Hg(II) on the response of the dual-sensing TOP10/pR-Pcad-Pmer-G toward Cd(II), the mixtures of 100 μM Cd(II) with 0–10 μM Hg(II) were added into the cultures of TOP10/pR-Pcad-Pmer-G during the lag phase. The fluorescent signals and bacterial density were measured after an 8-h induction at 37^oC.

Aliquots (100 μL) of bacterial culture were transferred into a 96-well microplate, and bacterial cell density was determined by measuring optical density at 600 nm using a microplate reader (BioTek Epoch, Winooski, VT, United States). Both green fluorescence and red fluorescence emitted from whole-cell biosensors were quantitated using a fluorescence spectrometer (Thermo, Waltham, MA, United States) as previously described (Hui et al., 2018c, 2021c). Briefly, bacterial cultures were diluted to 3 ml with purified water and pipetted into a low fluorescence background quartz cuvette. The intensity of emitted eGFP fluorescence was recorded at 507 nm with excitation at 488 nm, and the intensity of emitted mCherry fluorescence was recorded at 610 nm with excitation at 587 nm. The fluorescent signal is indicated as a fluorescence count value (unit = cnt) and finally calculated as relative fluorescent units (RFU) per OD₆₀₀.

Aliquots (50 μL) of bacterial cultures were spread onto glass slides. After being air-dried at room temperature, the slides were gently washed with purified water to remove the residues of culture ingredients. Bacterial cells were fixed on the glass slides, and then visible, and fluorescent images were taken under 400 × magnification with an Eclipse Ni fluorescence microscope (Nikon, Tokyo, Japan) equipped with a FITC and Texas Red filter as described previously (Hui et al., 2018a,b).

To investigate the differential activation of the divergent mer promoter, recombinant TOP10 harboring pR-Pmer-G was exposed to 0, 1.25, 2.5, 5, and 10 μM Hg(II). Bacterial density and double-color fluorescence were determined after 8-h incubation at 37^oC. As shown in Supplementary Figure 2A, the green fluorescence continuously increased with Hg(II) exposure at the concentrations range of 0–10 μM. However, the intensity of red fluorescence hardly changed. Thus it can be seen that the transcription of merR and its downstream mcherry in the opposite direction is not improved upon exposure to Hg(II). Similarly, to study the differential activation of the divergent cad promoter, TOP10/pG-Pcad-R was exposed to 0, 12.5, 25, 50, and 100 μM Cd(II). Bacterial density and double-color fluorescence were measured after 8-h incubation at 37^oC. As shown in Supplementary Figure 2B, significantly enhanced red fluorescence was observed with 0–100 μM Cd(II) exposure, and the change of the green fluorescence is unobvious. We can draw a conclusion that the transcription of cadR and its downstream egfp in the opposite direction is also not improved upon exposure to Cd(II). Single directional transcriptional activation contributes to the subsequent development of the dual-sensing construct.

To investigate the influence of insertion direction on the behaviors of biosensing constructs, TOP10/pPcad-R and TOP10/pR-Pcad in lag phase were exposed to 0, 12.5, 25, 50, and 100 μM Cd(II). The RFU was determined after 8-h incubation at 37^oC (Supplementary Figure 3). As expected, the red fluorescence in both groups increased with 0–100 μM Cd(II) induction. Notably, the responses of two biosensors exposed to the same Cd(II) concentration were similar.

Single-sensing TOP10/pPmer-G, single-sensing TOP10/pR-Pcad, and dual-sensing TOP10/pR-Pcad-Pmer-G were exposed to Zn(II), Pb(II), and Cd(II) at the concentration of 0, 5, 25, 125, or 250 μM, to Hg(II) at the concentration of 0, 5, or 25 μM. The RFU were determined after culturing at 37^oC for 8 h. Largely due to the high specificity of metalloregulator MerR, the green fluorescence, which is inducibly expressed from the merR-Pmer-egfp genetic cassette, selectively responded to Hg(II) in either a single-sensing construct (Figure 2A) or a dual-sensing construct (Figure 2C). A slightly enhanced response toward 250 μM Cd(II) was observed. The red fluorescent signal, which is induced from the cadR-Pcad-mcherry genetic cassette, strongly responded to Cd(II) in a dose-response relationship in either a single-sensing construct (Figure 2B) or a dual-sensing construct (Figure 2D). Furthermore, a strong red fluorescence responsive to Hg(II) was found in the single-sensing construct (Figure 2B), However, the red fluorescence responsive to Hg(II) was significantly decreased in the dual-sensing construct (Figure 2D).

Single-sensing TOP10/pPmer-G, single-sensing TOP10/pR-Pcad, and dual-sensing TOP10/pR-Pcad-Pmer-G were exposed to 0, 0.0024, 0.0049, 0.0098, 0.0195, 0.039, 0.078, 0.156, 0.3125, 0.625, 1.25, 2.5, 5, 10, 20, and 40 μM Hg(II). Bacterial density and double-color fluorescence were measured after culturing at 37^oC for 8 h. Bacterial densities of three whole-cell biosensors significantly decreased with above 10 μM Hg(II) exposure in two single-sensing biosensors (Figure 3A), probably due to the toxicity of high concentration of Hg(II). Overexpressed double-color fluorescent proteins made the Hg(II)-induced culture color green in TOP10/pPmer-G, red in TOP10/pR-Pcad, and yellow in TOP10/pR-Pcad-Pmer-G (Figure 3B). These colors, which could be easily recognized by the naked eye over a range of Hg(II) from 5 to 20 μM, inevitably interfered with the measurement of bacterial density at 600 nm. Due to the high absorbance of red color at 600 nm, the cultures with overexpressed mCherry always exhibited a high OD₆₀₀ value (Figure 3A). Single-sensing TOP10/pPmer-G with green fluorescence as the unique signal could detect as low as 0.0049 μM Hg(II), and dual-sensing TOP10/pR-Pcad-Pmer-G could detect as low as 0.0098 μM Hg(II) with green fluorescence as the output signal (Figure 3C). Furthermore, The detection limit for Hg(II) was improved to 0.3125 μM using both single-sensing TOP10/pR-Pcad and dual-sensing TOP10/pR-Pcad-Pmer-G with red fluorescence as the output signal (Figure 3F). The green fluorescence responsive to Hg(II) always increased upon exposure to Hg(II) below 10 μM, and the RFU of single-sensing TOP10/pPmer-G was significantly higher than that of dual-sensing TOP10/pR-Pcad-Pmer-G (Figure 3D). The concentration of Hg(II) and the RFU of eGFP in single-sensing TOP10/pPmer-G could be well-fitted by a non-linear regression relation in a narrow and low concentration range, and the regression equation and the coefficient of determination (R² = 0.99947) were shown in Figure 3E. The red fluorescence increased upon exposure to Hg(II) below 20 μM in signal-sensing TOP10/pR-Pcad. Then, the red fluorescence sharply decreased due to the high toxicity of 40 μM Hg(II) (Figure 3G). Therefore, the significant increased OD₆₀₀ value in TOP10/pR-Pcad culture exposed to 20 μM Hg(II) (Figure 3A) could be attributed to the high expression of mCherry. Interesting, the RFU of mCherry in dual-sensing TOP10/pR-Pcad-Pmer-G was not further increased upon exposure to Hg(II) upon 2.5 μM (Figure 3G). The Hg(II) concentration and the RFU of mCherry in single-sensing TOP10/pR-Pcad were also well-fitted by a non-linear regression relation in a wide and high concentration range. The regression equation and the coefficient of determination (R² = 0.99807) are shown in Figure 3H.

To confirm the double-color fluorescence of dual-sensing biosensor upon exposure to Hg(II), dual-sensing TOP10/pR-Pcad-Pmer-G was induced with 20 μM Hg(II), and TOP10/pR-Pcad-Pmer-G treated with 0 μM Hg(II) was used as the control group. After culturing at 37^oC for 8 h, biosensor cells were harvested and spread onto the slides. Bright-field and double-fluorescent field images were captured (Supplementary Figure 4). Strong double-color fluorescent signals were detected from biosensor cells treated with 20 μM Hg(II). Very weak double-color fluorescence was observed in the control group. Double-color fluorescent indication for Hg(II) was finally validated in the test.

Single-sensing TOP10/pR-Pcad and dual-sensing TOP10/pR-Pcad-Pmer-G were exposed to 0, 0.012, 0.024, 0.049, 0.098, 0.195, 0.39, 0.78, 1.56, 3.125, 6.25, 12.5, 25, 50, 100, 200, 400, and 800 μM Cd(II). Bacterial density and double-color fluorescence were analyzed after culturing at 37^oC for 8 h. Owing to significant cytotoxicity of Cd(II), bacterial densities of two whole-cell biosensors dramatically decreased upon exposure to above 400 μM Cd(II) exposure (Figure 4A). Overexpressed mCherry made the Cd(II)-inducible culture color red at 200 and 400 μM (Figure 4B), and it might enable an increased OD₆₀₀ when exposed to these two concentrations of Cd(II) (Figure 4A). Both single-sensing TOP10/pR-Pcad and dual-sensing TOP10/pR-Pcad-Pmer-G responded to as low as 0.098 μM Cd(II) (Figure 4C). The red fluorescent signals continuously increased when two biosensors were exposed to below 400 μM Cd(II), followed by a significant decrease probably due to the obvious cytotoxicity (Figure 4D). Compared with the red fluorescence of dual-sensing TOP10/pR-Pcad-Pmer-G, that of single-sensing TOP10/pR-Pcad upon exposure to above 200 μM Cd(II) was significantly stronger (Figure 4D). Importantly, the concentration of Cd(II) and the RFU of mCherry in both single-sensing TOP10/pR-Pcad and dual-sensing TOP10/pR-Pcad-Pmer-G were well-fitted by a non-linear regression relation in a wide and high concentration range, and the regression equations and the coefficients of determination were shown in Figure 4E. Interestingly, a strong green fluorescent signal responsive to above 200 μM Cd(II) was also observed in dual-sensing TOP10/pR-Pcad-Pmer-G (Figure 4F).

To further evaluate the performance of dual-sensing TOP10/pR-Pcad-Pmer-G upon exposure to Hg(II) and Cd(II) simultaneously, TOP10/pR-Pcad-Pmer-G was exposed to 2.5 μM Hg(II) in combination with 0, 25, 50, 100, 200, 300, or 400 μM Cd(II). Bacterial density (Figure 5A) and double-color fluorescence (Figure 5B) were measured after culturing at 37^oC for 8 h. Bacterial cell density showed a downward trend accompanied by an increased concentration of Cd(II) (Figure 5A). A significant rise in red fluorescence accompanied by 0–200 μM Cd(II) was observed. The intensity of red fluorescence remained stable at 200–400 μM Cd(II) and decreased at above 400 μM Cd(II) (Figure 5B). However, the green fluorescence just increased slightly at above 100 μM Cd(II), and it finally led to the significantly decreased ratios of eGFP/mCherry at concentrations of 0–200 μM Cd(II) (Figure 5B).

Then, TOP10/pR-Pcad-Pmer-G was exposed to 100 μM Cd(II) in combination with 0, 0.625, 1.25, 2.5, 5, or 10 μM Hg(II). After incubation at 37^oC for 8 h, bacterial density (Figure 6A) and double-color fluorescence (Figure 6B) were analyzed. Due to the increased cytotoxicity of heavy metal, a significant decline in bacterial cell density was observed at above 5 μM Hg(II) in combination with 100 μM Cd(II) (Figure 6A). A significant rise in green fluorescence accompanied with 0–10 μM Hg(II) was observed. However, just a slight rise in red fluorescence was found at above 1.25 μM Hg(II), and it undoubtedly led to the significantly enhanced ratios of eGFP/mCherry at concentrations of 0–10 μM Hg(II) (Figure 6B).