The E. coli strains and plasmids used in this study are listed in Table 1. Chemically competent E. coli Top 10 was used as a host for the cloning and propagation of plasmids. E. coli Top 10 and DH5α were used for induced expression of plasmid-coded red fluorescence protein mCherry, lacZ, and lacZα peptide. E. coli was grown in Lysogeny Broth (LB) (1% w/v tryptone, 0.5% w/v yeast extract, and 1% w/v NaCl) supplemented with 1% w/v glucose and 50 μg/mL ampicillin (Amp), as necessary (Bertani, 2004). All molecular biology reagents were obtained from TaKaRa (Dalian, China). 5-bromo-4-chloro-3-indolyl-β-D-galactoside (X-gal), Isopropyl-β-D-thiogalactopyranoside (IPTG), 2-Nitrophenyl-β-D-galactopyranoside (ONPG), L-arabinose, antibiotics, and lysozyme were purchased from Sangon Biotech (Shanghai, China). Tryptone and yeast extract were obtained from OXOID (Basingstoke, United Kingdom). All chemicals were purchased from Sigma-Aldrich (Indianapolis, United States). All oligonucleotides and some fusion genes were synthesized by Sangon Biotech (Shanghai, China).

For cloning purposes, all PCR reactions were carried out for 25 cycles of 2 min at 95°C, 1 min at 50–60°C, and 1–2 min at 72°C. All primers used in this study are listed in Table 2.

A cassette containing the lac promoter, a lacZα open reading frame (ORF), and rrnB termination region was synthesized by Sangon Biotech (Shanghai, China). The synthesized fragment was then cloned as a NdeI/EcoRI fragment into pBR322, previously digested with the same enzymes. The resulting plasmid was named pPlac-lacZα. Another fusion cassette containing the lac promoter, a SalI restriction enzyme site, a full-length β-galactosidase lacZ ORF, and rrnB terminator was also synthesized. The resulting Plac-lacZ-rrnB terminator gene fusion cassette was then digested and inserted in pBR322 via NdeI and EcoRI sites, generating Plac-lacZ.

The pPlac-RFP was constructed through a PCR overlap extension method (Miller et al., 2000). First, the lac promoter region was amplified from pPlac-lacZα using primers F-Plac and R-Plac-RFP, and mCherry was amplified from pT-RFP using primers F-Plac-RFP and R-RFP-Ter. Second, both of the amplified products from the above reactions were added to a second PCR reaction using primers F-Plac and R-RFP-Ter, and the resulting fragment was the Plac-mCherry cassette. Finally, the rrnB terminator was amplified from pPlac-lacZα using primers F-RFP-Ter and R-Ter. Then the amplified product and the Plac-mCherry cassette were added to a second PCR reaction using primers F-Plac and R-Ter. The resulting Plac-mCherry-rrnB terminator gene fusion cassette and pBR322 were then digested with NdeI and EcoRI, and ligated together to generate the plasmid pPlac-RFP.

The cassette containing the sequence encoding araC and an ara promoter region was amplified from pBAD vector using primers F-Para and R-Para. This fragment was then cloned as a NdeI and SalI fragment into pPlac-lacZ, previously digested with the same enzymes. The resulting plasmid was named pPara-lacZ. The cassette containing the gene of mCherry and rrnB terminator region was amplified from pPlac-RFP using primers F-RFP and R-Ter. The amplified fragment was then cloned as a SalI and EcoRI fragment into pPara-lacZ to generate the plasmid pPara-RFP. The cassette containing the gene of lacZα and rrnB terminator region was amplified from pPlac-lacZα using primers F-lacZα and R-Ter. The resulting PCR fragment was digested and inserted into pPara-lacZ via SalI and EcoRI sites, generating pPara-lacZα.

The strategy used for detection of lac and ara promoter activity with different reporter systems is summarized in Figure 1. A 2063 bp NdeI-EcoRI fragment of plasmid pBR322, which only retains the origin of replication of pMB1, and the gene bla encoding the ampicillin resistance (Amp^R) protein, was chosen to be the cloning vector backbone. All promoter reporter vectors share a common plasmid pBR322 backbone, but contain different promoter reporter cassettes. The target promoter, the ribosome binding site (RBS), a reporter encoding sequence and the terminator (derived from E. coli rrnB operon) (Brosius et al., 1981) were first fused. The resultant fusion cassette and the clone vector backbone were then assembled with restriction enzymes NdeI and EcoRI to create the reporter vectors.

The pbr operon, originating from Cupriavidus metallidurans strain CH34, is a unique lead resistance operon (Borremans et al., 2001; Vandamme and Coenye, 2004; Hui et al., 2018a). Based on the natural pbr operon, the lead bacterial biosensors have been successfully developed, and the genetic elements of the lead biosensor constructs consist of the transcriptional factor PbrR gene, together with a divergent pbr promoter, and a promoterless reporter fluorescent protein gene (Wei et al., 2014; Bereza-Malcolm et al., 2016). To test the detection sensitivity of pbr promoter activity, pPpbr-RFP, pPpbr-lacZα, and pPpbr-lacZ were constructed in this study.

The cassette containing the sequence encoding the transcriptor PbrR and the bidirectional pbr promoter region was synthesized by Sangon Biotech (Shanghai, China), and the synthesized fragment (520 bp) was cloned into pUCm-T to generate pT-Ppbr. The lacZα and mCherry pbr promoter reporter cassettes were constructed as follows: A lacZα reporter element and a mCherry reporter element were amplified from pPlac-lacZα, and pPlac-RFP, respectively, and fused to the genetic element containing pbrR and pbr promoter by a PCR overlap extension method. Briefly, a lacZα reporter element was amplified with primers F-Ppbr-lacZα and R-Ter, and a mCherry reporter element was amplified with primers F-Ppbr-RFP and R-Ter. The genetic element containing pbrR and pbr promoter was amplified from pT-Ppbr using primers F-Ppbr and either R-Ppbr-lacZα or R-Ppbr-RFP. Both of the amplified products from pPlac-lacZα and pT-Ppbr (using primer R-Ppbr-lacZα) were added to a second PCR reaction containing primers F-Ppbr and R-Ter to generate the Ppbr-lacZα gene fusion cassette. Similarly, the second round of PCR with primers F-Ppbr and R-Ter was performed after mixing the amplified product from pT-Ppbr (using primer R-Ppbr-RFP) with the amplified product from pPlac-RFP to generate the Ppbr-RFP gene fusion cassette. The two Ppbr reporter cassettes above and pBR322 were finally digested with NdeI and EcoRI, and ligated together to generate pPpbr-lacZα, and pPpbr-RFP, respectively. The cassette containing the sequence encoding PbrR and a pbr promoter region was amplified from the pT-Ppbr vector using primers F-Ppbr and R-Ppbr. This fragment was then cloned as a NdeI and SalI fragment into pPlac-lacZ, previously digested with the same enzymes. The resulting plasmid was named pPpbr-lacZ.

The strategy and potential mechanism involved in three reporter systems are shown in Figure 2. β-galactosidase α-complementation can be achieved by assembling the lead(II) inductive lacZα derived from the plasmid with the IPTG inductive lacZM15 derived from the host cell genome.

Escherichia coli hosts were transformed with recombinant vectors using a CaCl₂-mediated transformation method (Hui et al., 2018b). The transformed E. coli cells were spread onto LB agar plates containing 50 μg/mL ampicillin, and cultured overnight at 37°C. A single colony picked from an agar plate was used to inoculate 3 mL of LB medium supplemented with 50 μg/mL ampicillin in a 15 mL Bio-Reaction tube (Jet, Guangzhou, China), and cultured for 12 h in a 37°C shaking incubator at 200 rpm.

For IPTG-induced reporter genes expression, recombinant E. coli harboring the lac promoter reporter vectors were cultured in LB medium for 12 h, and diluted to an OD₆₀₀ of 0.01 in 12 mL fresh LB medium supplemented with 1% glucose and 50 μg/mL ampicillin in a 50 mL Bio-Reaction tube. The culture was grown at 37°C for 3 h with rotation at 220 rpm, and the bacteria reached log phase with an optical density 0.4 at 600 nm. The cultures were then induced with 0–1.0 mM IPTG and incubated at 37°C with shaking at 220 rpm. Induced cultures were sampled and evaluated for the expression of reporter genes at regular intervals.

For arabinose-induced reporter genes expression, recombinant E. coli harboring the ara promoter reporter vectors were cultured in LB medium for 12 h, and diluted to an OD₆₀₀ of 0.01 in 12 mL fresh LB medium supplemented with 50 μg/mL ampicillin plus 0.1 mM IPTG in a 50 mL Bio-Reaction tube. The cultures were grown at 37°C for 3 h with rotation at 220 rpm, and then induced with varying concentrations of arabinose. Induced cultures were incubated at 37°C with shaking at 220 rpm, and tested for the expression of reporter genes at regular intervals.

For Pb(II)-induced reporter genes expression, recombinant E. coli harboring the pbr promoter reporter vectors were cultured in LB medium for 12 h, and diluted to an OD₆₀₀ of 0.01 in 12 mL fresh LB medium supplemented with 50 μg/mL ampicillin in a 50 mL Bio-Reaction tube. The cultures were grown at 37°C for 3 h with rotation at 220 rpm, and then induced with varying concentrations of Pb(II) plus 0.1 mM IPTG. Induced cultures were incubated for 4 h at 37°C with shaking at 220 rpm, sampled and subjected to reporter gene assays.

An enzymatic assay of β-galactosidase activity using cell lysate was performed and modified as previously described (Tomizawa et al., 2016). In brief, E. coli cells were collected, washed twice with 50 mM phosphate buffer saline (PBS, pH 7.0), and resuspended in 50 mM PBS. After the OD₆₀₀ of the resuspended cells was measured, an equal volume of assay solution (1 mg/mL lysozyme, 2 mM X-gal, 200 mM NaCl, 50 mM PBS) was added, vortexed for 2 min, and incubated at 37°C for 30 min. The supernatant was obtained by centrifuging at 3500 rpm for 5 min, and then the absorbance was measured at 630 nm using an iMark microplate reader (Bio-Rad, United States). The background value was obtained from the control assays with uninduced cell lysate, and the optical density at 630 nm was then normalized by the OD₆₀₀ value of cell suspensions.

The fluorescence intensity of mCherry produced in E. coli was measured as previously described (Hui et al., 2018c). Briefly, E. coli cells were collected, washed, and resuspended in 50 mM PBS. A 3-mL aliquot of E. coli suspension or diluent was added to 1-cm low fluorescence background quartz cuvette. To test the fluorescence intensity of induced mCherry, the excitation wavelength was set at 587 nm and the intensity of emitted fluorescence of mCherry at 610 nm was recorded with Lumina fluorescence spectrometer (Thermo Fisher Scientific, United States). The fluorescence intensity was normalized by dividing the fluorescence intensity at the emission wavelength 610 nm by the OD₆₀₀ value of the same sample. The background value was obtained from the control assays with uninduced cell suspensions.

Single colonies were picked from each construct and inoculated in 3 mL LB medium with 50 μg/mL ampicillin at 37°C for 6 h. The precultured recombinant E. coli was plated on LB plate containing optimal concentration of inducer for assay, or no inducer for the control, cultured at 37°C for 18 h. The LB agar was supplemented with 0.6 mM IPTG to induce the transcription of lac promoter, 0.002% arabinose plus 0.1 mM IPTG to induce the transcription of ara promoter, and 20 μM Pb(II) plus 0.1 mM IPTG to induce the transcription of pbr promoter. To examine the expression of reporter lacZα and LacZ, extra substrate 0.4% X-gal or 0.4% ONPG was also added to LB agar.

In order to compare the transcriptional and translational levels of the novel lacZ α-complementation reporter system, the full-length β-galactosidase, and the commonly used fluorescent protein reporter system, a monocistronic lacZα, lacZ, and mCherry reporter vectors based on an artificial lac operon were assembled, and named pPlac-lacZα, pPlac-lacZ, and pPlac-RFP, respectively.

First, we investigated the response profiles of recombinant E. coli Top10 harboring pPlac-lacZα, pPlac-lacZ, and pPlac-RFP in response to varying concentrations of IPTG. After bacterial cells in logarithmic growth were exposed to 0–1.0 mM IPTG for 8 h at 37°C, β-galactosidase activity was assayed in Top10/pPlac-lacZα and Top10/pPlac-lacZ cultures, and mCherry fluorescent intensity was determined in Top10/pPlac-RFP cultures. The results are shown in Figure 3A. Top10/pPlac-RFP was found to respond to the lowest concentration of IPTG at 0.1 mM. However, even when the concentration of IPTG reached as low as 0.001 mM IPTG, β-galactosidase activity was still detected in both Top10/pPlac-lacZα and Top10/pPlac-lacZ cultures. With the increase of IPTG concentration, although significantly higher response levels of three reporters were observed, there was no significant difference of β-galactosidase activities in Top10/pPlac-lacZα and Top10/pPlac-lacZ with 0.001–0.2 mM IPTG induction. The highest response levels of Top10/pPlac-lacZα and Top10/pPlac-lacZ were obtained at about 0.2 mM IPTG, and 0.4 mM IPTG, respectively. The highest response level of Top10/pPlac-RFP was obtained at about 0.6 mM IPTG. The preferred concentration of IPTG was finally determined to be 0.6 mM for the following response time assay.

Second, response times of Top10/pPlac-lacZα, Top10/pPlac-lacZ, and Top10/pPlac-RFP were analyzed following exposure to 0.6 mM IPTG. The results are shown in Figure 3B. Cultures were sampled at consecutive time intervals after IPTG exposure. An approximate 3-h delay in the mCherry fluorescent signal was observed in Top10/pPlac-RFP cultures, and the highest fluorescent intensity was obtained at around 6 h. The response times of Top10/pPlac-lacZα and Top10/pPlac-lacZ were as low as 0.5 h. β-galactosidase activity derived from both lacZα and wild type lacZ increased with prolongation of the induction time up to 4 h, at which point the enzyme activity reached a maximum. Then, the enzyme activity of Top10/pPlac-lacZα decreased gradually, and remained at approximately 88% of the maximum at 10 h.

The production of active β-galactosidase depends on the expression of both α-donor lacZα peptide and α-acceptor lacZM15 protein (Mogalisetti and Walt, 2015). The inductive expression profile of lacZα encoded in a reporter vector might be variable in different hosts. To further validate the performance of lacZα reporter system in different hosts, we examined the response levels of Top10/pPlac-lacZα and DH5α/pPlac-lacZα in response to 0.001, 0.01, 0.1, and 1.0 mM IPTG, respectively. IPTG at varying concentrations was added to an exponential phase bacterial culture and incubated for 4 h at 37°C. Interestingly, Top10/pPlac-lacZα was found to produce significantly higher β-galactosidase activity at all IPTG concentrations examined, in comparison to DH5α/pPlac-lacZα (Supplementary Figure S2). It is well known that the expression of recombinant protein is determined by host genetic background, plasmid copy number, inducer concentration, culture conditions, and so on (Hortsch and Weuster-Botz, 2011; Mahalik et al., 2014). E. coli Top10 was then chosen as a preferred host for the following tests.

Furthermore, β-galactosidase activity was not elevated above 0.1 mM IPTG induction, and there was no significant difference in the growth curves of Top10/pPlac-lacZα and DH5α/pPlac-lacZα with 0–0.1 mM IPTG induction (data not shown). Thus, 0.1 mM IPTG was used for the induced expression of lacZM15 in the following promoter activity assays.

To further determine the potential of lacZα reporter system in detecting promoter activity using a promoter other than the lac promoter, a monocistronic lacZα, lacZ, and mCherry reporter constructs based on an artificial ara operon were assembled.

The response profiles of Top10/pPara-lacZα, Top10/pPara-lacZ, and Top10/pPara-RFP in response to varying concentrations of inducer arabinose was first determined. After recombinant E. coli in logarithmic growth were exposed to 0–0.02% arabinose for 8 h at 37°C, β-galactosidase activity and mCherry fluorescent intensity were assayed. As shown in Figure 4A, Top10/pPara-lacZα was demonstrated to respond to as low as 0.0001% arabinose. With the increase of arabinose concentration, significantly increased response of lacZα production-inducing β-galactosidase activity was achieved. The highest response level of Top10/pPara-lacZα was obtained at 0.0006% arabinose. However, Top10/pPara-lacZ and was found to respond to the lowest concentration of arabinose at 0.0002% after 8-h induction. Although the response level of Top10/pPara-lacZ always increased with the increase of arabinose concentration, the response level of full-length lacZ reporter was still significantly lower than that of lacZα reporter with 0.0001–0.002% arabinose induction. Top10/pPara-RFP was found to respond to the lowest concentration of arabinose at 0.0006% after 8-h induction, and the highest response level of Top10/pPara-RFP was obtained at 0.001% arabinose.

Response times of Top10/pPara-lacZα, Top10/pPara-lacZ, and Top10/pPara-RFP were analyzed following induction with 0.002% arabinose. The results are shown in Figure 4B. An approximate 2-h delay in the mCherry fluorescent signal was observed in Top10/pPara-RFP, and the highest fluorescent intensity was obtained at 6 h. An approximate 1-h delay in wild type lacZ activity was observed in Top10/pPara-lacZ, and the lacZ activity increased with prolongation of the induction time up to 10 h. Interestingly, nearly no time delay was observed in lacZα production-inducing β-galactosidase activity. In addition, the response level of lacZα reporter increased with prolongation of the induction time up to 3 h, at which point the enzyme activity had reached a maximum. Importantly, the response level of lacZα reporter was always higher than that of full-length lacZ reporter within a 6-h induction, and the biggest difference between the two groups occurred within a 4-h induction.

The inducible pattern of lacZα reporter is different from a standard lacZ reporter, in which a truncated lacZα is substituted for a wild type lacZ as a reporter. Importantly, the expression of lacZM15 is independent of the inducible expression of lacZα. Through α-complementation, lacZα peptide can be used as a reporter for the target promoter, because α-acceptor lacZM15 is pre-expressed in the host cell in advance. Obviously, induced expression of a small peptide lacZα would be easier and faster in E. coli cells than induced expression of a large-sized lacZ.

The application of lacZα reporter system is limited in lacZM15-producing bacteria, and this is its only disadvantage. When a moderately strong ara promoter was placed in front of three reporter cassettes, the results above indicate that the most sensitive response was achieved in a lacZα reporter vector. To examine whether the lacZα-derived report system can be used to detect low-level transcriptional signals from a weak promoter other than ara promoter, a 520 bp fragment containing pbrR and pbr promoter was placed in front of promoterless mCherry, lacZ, and lacZα reporter cassettes. The pbr promoter activity was then evaluated by assays of both mCherry and β-galactosidase activity. E. coli transformed with Ppbr-RFP, Ppbr-lacZ, and Ppbr-lacZα showed increased signals in response to elevated concentrations of Pb(II). The results are shown in Figure 5.

Recombinant Top10/pPpbr-RFP, Top10/pPpbr-lacZα, and Top10/pPpbr-lacZ in logarithmic growth were exposed to varying concentrations of Pb(II) plus 0.1 mM IPTG for 4 h at 37°C. Top10/pPpbr-RFP, a lead biosensor with a single mCherry reporter system, detected 15 μM Pb(II) (Figure 5A). This result was basically consistent with previous findings (Bereza-Malcolm et al., 2016). However, the lacZα peptide, like full-length lacZ, was demonstrated to be a much more sensitive reporter to detect low transcriptional activity. Both Top10/pPpbr-lacZα and Top10/pPpbr-lacZ were demonstrated to respond to concentrations as low as 3 μM Pb(II) (Figure 5B). It is worth mentioning that the responses from both lacZα and lacZ report systems were associated with Pb(II) exposure in a dose-response manner, and lacZα production-inducing β-galactosidase activity was significant higher than wild type β-galactosidase activity with 3–10 μM Pb(II) induction. However, no significantly higher response from the mCherry reporter system was observed with the increase of Pb(II) in the study. These results suggest that even a weak promoter with very poor transcriptional activity may still be detected with a lacZα-derived reporter system.

The activity of a target promoter in response to an inducer can be quantified in the liquid media. Moreover, like the expression of mCherry and wild type β-galactosidase, the activation of β-galactosidase in lacZα-derived reporter systems could also be clearly observed with the naked eye in a plate assay (Figure 6). Both X-gal and ONPG could be used as substrates for the chromogenic reaction to indicate the production of active β-galactosidase.