Sensitive and Specific Whole-Cell Biosensor for Arsenic Detection

Arsenic poisoning is a severe public health issue. Rapid and simple methods for the sensitive and specific monitoring of arsenic concentration in drinking water are needed. In this study, we designed an arsenic WCB with a positive feedback amplifier. It is highly sensitive and able to detect arsenic below the WHO limit level. In addition, it also significantly improves the specificity of the biosensor toward arsenic, giving a signal that is about 10 to 20 times stronger in response to As(III) than to other metals. This work not only provides simple but effective arsenic biosensors but also demonstrates the importance of genetic engineering, particularly the use of positive feedback amplifiers, in designing WCBs.

ABSTRACT

Whole-cell biosensors (WCBs) have been designed to detect As(III), but most suffer from poor sensitivity and specificity. In this paper, we developed an arsenic WCB with a positive feedback amplifier in Escherichia coli DH5α. The output signal from the reporter mCherry was significantly enhanced by the positive feedback amplifier. The sensitivity of the WCB with positive feedback is about 1 order of magnitude higher than that without positive feedback when evaluated using a half-saturation As(III) concentration. The minimum detection limit for As(III) was reduced by 1 order of magnitude to 0.1 µM, lower than the World Health Organization standard for the arsenic level in drinking water, 0.01 mg/liter or 0.13 µM. Due to the amplification of the output signal, the WCB was able to give detectable signals within a shorter period, and a fast response is essential for in situ operations. Moreover, the WCB with the positive feedback amplifier showed exceptionally high specificity toward As(III) when compared with other metal ions. Collectively, the designed positive feedback amplifier WCB meets the requirements for As(III) detection with high sensitivity and specificity. This work also demonstrates the importance of genetic circuit engineering in designing WCBs, and the use of genetic positive feedback amplifiers is a good strategy to improve the performance of WCBs.

IMPORTANCE Arsenic poisoning is a severe public health issue. Rapid and simple methods for the sensitive and specific monitoring of arsenic concentration in drinking water are needed. In this study, we designed an arsenic WCB with a positive feedback amplifier. It is highly sensitive and able to detect arsenic below the WHO limit level. In addition, it also significantly improves the specificity of the biosensor toward arsenic, giving a signal that is about 10 to 20 times stronger in response to As(III) than to other metals. This work not only provides simple but effective arsenic biosensors but also demonstrates the importance of genetic engineering, particularly the use of positive feedback amplifiers, in designing WCBs.

INTRODUCTION

Arsenic (As)-contaminated groundwater, occurring from mining or agriculture or natural contamination due to the abundance of arsenic in the Earth’s crust, is a serious global health issue. Long-term exposure to arsenic can result in various diseases including cancers (1, 2). It is estimated that over 100 million people worldwide may be at risk from consuming water contaminated with arsenic (3). The World Health Organization (WHO) has recommended 0.01 mg/liter (0.13 µM) as a safe permissible level for arsenic in drinking water (4), and the Food and Agriculture Organization (FAO) has set a maximum contamination level (MCL) for arsenic of 0.01 mg/liter in irrigation water (5). Due to its toxicity and strict arsenic standards for drinking water, cost-effective and sensitive environmental monitoring tools to detect arsenic are needed.

To date, many methods have been reported to detect arsenic at low concentrations, such as chemiluminescent immunoassay (6), inductively coupled plasma optical emission spectrometry (ICP-OES) (7), and atomic absorption spectrometry (AAS) (8). However, these methods often require complicated and expensive instruments and trained professionals to pretreat and analyze samples, making them hard to use in situ (9, 10). To overcome these limitations, biosensors using enzymes, antibodies, and microorganism cells have garnered interest for use in the detection of arsenic in drinking water (11, 12).

Recently, whole-cell biosensors (WCBs) have been extensively studied for the specific and sensitive detection of toxic heavy metal ions. Often, the regulatory elements from a heavy metal resistance operon, including the transcriptional regulator and its cognate promoter, are coupled to a reporter gene such as fluorescence, luminescence, or enzyme assays so that the signal strength from the reporter is correlated to the concentration of the heavy metal to be detected. The key to developing a sensitive and specific WCB is to identify the regulatory elements and then optimize performance by engineering the regulatory elements or the genetic circuit. A relatively well-studied arsenic resistance operon is the one found in Escherichia coli, which contains arsR (transcriptional regulator), arsB (arsenite permease), and arsC (arsenate reductase) (13, 14). When arsenic is absent, the transcription regulator ArsR binds to the ArsR-binding site (ABS) within the ars promoter and blocks transcription. Once arsenic is present, it binds to ArsR and changes the local structure of the promoter to activate the transcription of the ars genes and clear arsenic in the cell (14–16). The arsR regulator and the promoter of this operon have been used to construct arsenic WCBs in various microorganism hosts (11, 17, 18). However, low sensitivity and specificity are major issues when using them for arsenic detection below the WHO recommended level (19–21).

In this study, we used E. coli as the host to construct arsenic WCBs since it naturally contains the ars operon. Genetic circuit engineering was done to improve the performance of the WCB. Positive feedback is common in nature and well known for signal amplification (22). It has been used to improve the sensitivity of WCBs in response to various analytes, including antibiotics, amino acids, and heavy metals (23). This work introduced a positive feedback loop using the LuxR autoregulatory elements to arsenic WCBs for the first time to improve sensitivity and specificity. The comparison of the designs with and without the positive feedback amplifier in this work provides useful insights for the development of WCBs in the future.

RESULTS

Design and construction of the biosensors.

As shown in Fig. 1, two arsenic WCBs were constructed in E. coli DH5α. The first one simply coupled the arsenic-inducible promoter (P_ars) and its regulatory gene (arsR) with the reporter gene mCherry. The signal from mCherry is directly correlated to the concentration of the inducer arsenic. No positive feedback circuit was involved. In the second one (Fig. 1B), the transcriptional activator, a variant of luxR, was used to replace mCherry, and it was regulated by the arsR-P_ars circuit, while mCherry together with a second luxR was placed under the promoter P_luxI, which was activated by LuxR. When arsenic is present, it activates the expression of LuxR in the first plasmid, which turns on the expression of mCherry and LuxR from the second plasmid. The second LuxR activates its own expression as well as that of mCherry and forms a positive feedback loop to enhance the output signal from mCherry in the second plasmid. These two plasmids work together as the arsenic WCB with the positive feedback amplifier.

FIG 1.

Schematic of the arsenic WCBs with positive feedback (B) and without (A). (A) The typical arsenic WCB consists of the ArsR-regulated promoter P_ars, the regulator arsR, and the reporter gene mCherry. (B) The positive feedback WCB consists of the arsR-P_ars regulatory circuit and a positive feedback amplifier where LuxR produced in response to arsenite activates the expression of mCherry and LuxR from the P_luxI promoter. The LuxR from the P_luxI promoter activates its own expression and forms a positive feedback loop.

Growth curve of the WCB strains.

To understand the toxic effects of arsenic on the engineered strains, the growth curves of DH5α/pCDF-As-mCherry and DH5α/pCDF-As-luxR+pGN68-mCherry at different concentrations of arsenic were measured. Arsenic at a final concentration of 0, 0.1, 1, 10, 100, 200, 300, 400, 500, or 600 µM was added to the subculture, and optical density at 600 nm (OD₆₀₀) was measured every hour for 10 h. As shown in Fig. 2, for both strains, no significant effect on the growth was observed when the concentration of As(III) was below 10 μM. The bacteria entered the logarithmic phase after incubation for 2 h and the stationary phase after about 6 h. When the concentration of As(III) was at or above 100 μM, cells grew much slower. The arsenic toxic effect was more obvious and cells barely grew when the As(III) concentration was above 200 μM. Therefore, 0 to 200 μM As(III) was used to obtain the arsenic dose-response curve in the next section.

FIG 2.

Growth curves of the two WCBs at different concentrations of arsenite. (A) WCB without positive feedback; (B) WCB with positive feedback.

Arsenic sensitivity and specificity of WCBs.

The expression of the reporter mCherry from these WCBs accumulates along with time. Not only As(III) concentration but also the exposure time affect the output signal of the arsenic WCBs (24–26). Therefore, the time-response curves were measured before examining the sensitivity and specificity of the two arsenic WCBs.

Time-dependent response.

The response time of a WCB is an essential factor for practical application. In addition, a potential issue using a genetic amplifier in WCBs is that the basal-level expression, either from the sensing module or the amplifying module, may be self-reinforcing and cause a high level of false-positive signal over time. Therefore, the time-dependent responses of the two WCBs were monitored for 10 h after adding 0 μM, 0.01 μM, 0.1 μM, or 10 μM As(III). As shown in Fig. 3, the background signal without As(III) slightly increased when the incubation time was above 6 h for both WCBs. After incubation for 10 h without arsenic, the fluorescence signal from the positive feedback biosensor was increased by 5.4 times, while the nonpositive feedback biosensor showed a 4.8-times increase. Therefore, the basal level expression was comparable for the biosensors with and without the positive feedback amplifier. A false signal from the amplification of potential leaky expression was not noticed.

FIG 3.

Time-dependent response of arsenic biosensors with (A) and without (B) positive feedback. Biosensor cells were grown for 10 h at 0 μM, 0.01 μM, 0.1 μM, and 10 μM As(III). Statistical significance was shown as follows: *, P < 0.01; **, P < 0.001.

The response of the positive feedback amplifier biosensor to As(III) was faster, and the signal was much higher than that of the one without positive feedback. After As(III) was added at a concentration of 0.1 μM for 6 h, the WCB with the positive feedback amplifier exhibited fluorescence that was about 3-fold stronger than that without the positive feedback. In addition, the output signal of the positive feedback amplifier biosensor exposed to 10 μM As(III) for 4 h was 11 times higher than that of the biosensor without positive feedback.

Dose-dependent response to arsenite.

The amplification effect of the positive feedback loop with different initial concentrations of arsenic was analyzed. The two WCBs were compared after exposure to As(III) at 0 to 200 μM for 9 h at 37°C.

Both WCBs displayed a similar dose-dependent pattern with the fluorescence intensity positively correlated with the concentrations of As(III) (Fig. 4). Also, it is noted that the sensitivity of the WCB with the positive feedback amplifier was higher by approximately 1 order of magnitude compared to that of the WCB without positive feedback. The half-saturation As(III) concentration that gave half of the maximum mCherry fluorescence intensity for the WCB with positive feedback was about 0.5 to 1 μM, while the half-saturation As(III) concentration for the WCB without positive feedback is about 10 to 50 μM. Moreover, the WCB with positive feedback significantly amplified the output signal, about 2.5 to 5.5 times of that from the WCB without positive feedback when As(III) went from 0.1 to 100 μM. The expression of the mCherry gene was noticeable when As(III) was added at a concentration as low as 0.1 µM (P < 0.01) for the WCB with positive feedback and 0.5 μM (P < 0.001) for the one without positive feedback. The detection limit of the WCB with positive feedback is lower than the WHO drinking water standards and potentially could be developed as arsenic biosensors in real application. These results suggested that, compared with the WCB without positive feedback, the one with a positive feedback amplifier functions well in enhancing the fluorescence intensity, increasing the detection range, and improving sensitivity.

FIG 4.

Dose-response curves of arsenic biosensors with (red solid squares) and without (black solid squares) the positive feedback amplifier. Biosensors were grown for 9 h at different arsenite concentrations from 0 μM to 200 μM. The right y axis indicates that the maximum FIR value is set to 1. Statistical significance was shown as follows: *, P < 0.01; **, P < 0.001.

Specificity of the WCBs.

In addition to sensitivity and strength of output signal, specificity toward arsenic is also an important factor to evaluate the WCB. Both WCBs were exposed to NaAsO₂, Pb(NO₃)₂, ZnCl₂, CuCl₂, and CdCl₂, at a final concentration of 0.01 µM, 0.1 µM, 1 µM, or 10 µM, and the fluorescence of mCherry was measured after 8 h. As shown in Fig. 5, compared to As(III), the response of the WCB with positive feedback to other metals was negligible at either 1 µM or 10 µM, which is less than 2% or 10% of that to As(III). At 0.1 µM, the signal from other metals was about 10% to 60% of that from As(III). However, the fluorescence response of the WCB without positive feedback to other metals was about 37% to 71% at 0.1 µM, 20% to 30% at 1 µM, and 15% at 10 µM of that from As(III). By introducing the positive feedback amplifier into the arsenic WCB, the output signal was enhanced so much that the specificity of the WCB toward arsenic was also significantly increased, which is remarkable as no other designs have been reported to be able to improve the specificity of a WCB through circuit engineering.

FIG 5.

Arsenic specificity of the biosensors with and without the positive feedback amplifier. Fluorescence intensity of the biosensors was measured after exposure to various metals at concentrations of 0.01 μM, 0.1 μM, 1 µM, and 10 µM for 8 h.

DISCUSSION

WCBs have been extensively explored in order to detect toxic heavy metals and metalloids in environments, including cadmium, lead, mercury, and arsenic (27–30). Although many WCBs have been constructed and studied for the detection of arsenic (31–33), most of these WCBs have not been used for environmental monitoring because of the high requirements of sensitivity and specificity (34–37). K. de Mora and colleagues have reported a sensitive biosensor for arsenic in which pH is an input signal, and a color change is the output signal (34). It can detect less than 10 µg/liter (10 ppb) of As(III) with static overnight incubation. Nevertheless, the incubation time is too long for practical applications. L. A. Pola-Lopez et al. developed a new vector where the RNA polymerase of bacteriophage T7 was used as an amplifier with green fluorescent protein (GFP) as the reporter (31). The detection range of this biosensor was between 5 and 140 µg/liter. As(III) concentrations below WHO standards can be detected. However, in contrast to our designs, the amplifier did not translate into an improvement of biosensor performance, and the amplification effect was not evaluated using the unamplified biosensor as a control.

Genetic amplifiers have been used to construct WCBs to intensify the output signal and increase sensitivity to analytes (22, 23), but the effect of a positive feedback loop on increasing the sensitivity of the circuit is different from case to case, depending on the genetic context and the regulatory element it is coupled with. The LuxR positive feedback loop has been used for designing various WCBs, either increasing the output signal or improving the sensitivity, but has not been reported for detecting arsenic. Our work showed that when coupled with the arsR regulatory circuit, the LuxR positive feedback circuit not only increased its sensitivity but also improved its selectivity toward As(III).

One concern about incorporating a positive feedback circuit in WCBs in general is the high noise level and the false-positive signal from amplification of the basal-level expression from leaky promoters. K. Bansal et al. applied the positive feedback amplifier to modulate the expression kinetics of membrane proteins (38). This showed a statistically significant increase in the rate of production of the bd oxidase membrane protein. In addition, the positive feedback plays a role in implementing bistability, which is conventionally named high/low or ON/OFF in steady-state levels of gene expression. In other words, the gene expression levels of the biosensor are determined by the initial concentration of the inducer, and the level of the initial input inducer can cause changes in gene expression between the two steady-state levels. Single-cell measurements showed that whether positive feedback amplifiers increased cell noise compared with nonpositive feedback controls depends on the activity of LuxR proteins. Similar noise levels were observed for both the positive feedback and nonpositive feedback systems. In contrast, the increased noise at higher inducer concentrations may be a result of increased system burden by the reduced growth rates of the cells. In our work, the variation of fluorescence from the WCB with positive feedback was actually slightly lower than that from the WCB without positive feedback. The time-response curve in the absence of As(III) showed that the output signal accumulated to roughly the same extent for the two WCBs. Amplification of leaky signals was not observed.

The arsenic biosensor with the positive feedback amplifier that we designed has a broader detection range and a lower detection limit than the one without positive feedback because the positive feedback loop can amplify the output signal, which also increases sensitivity, as the accumulation of fluorescent proteins occurs at a low concentration of inducer arsenic. The positive feedback biosensor constructed in this study can detect arsenic as low as 0.1 μM and is more sensitive than the biosensors constructed thus far (39, 40). The half-saturation As(III) concentration of the biosensor with positive feedback is about 1 order of magnitude lower than that without positive feedback. An amplifier has been applied to detect cadmium and lead. It used T7 RNA polymerases to modulate multiple circuits by decoupling the transcription from the host and enhancing the expression (41, 42). It can reduce the detection limit for cadmium ions but has little effect on the minimum detection limit of lead ions. In addition to improving the sensitivity of biosensors, the positive feedback amplification system demonstrated improved specificity toward arsenite, as it significantly amplified the signal in response to arsenite but only marginally increased the signal in response to other metals. So overall, it showed a dramatic difference in the fluorescence in response to arsenite and other metals. Many arsenite WCBs have specificity issue (39, 43). Generally, the specificity may be improved by protein engineering to modify the interaction between the regulator and the inducer metal ions. It is surprising that the positive feedback loop introduced to the arsenite WCB in this study differentially enhanced the output signal in response to arsenite and other metals. The arsenite-induced signal was 10- to 40-fold higher than that of other metals, which is sufficient to detect As(III) in the presence of other metals.

The minimum detection limit of heavy metal biosensors often refers to the final concentration of the metal in the medium under optimal culture conditions (44). In this work, the same definition was used to describe the detection limit of the biosensors so that it can be compared with the work of other groups. A concern for the practical application is the difference between the real detection limit and the final concentration since the analyte is diluted when adding the sample water to the culture. It may not be detectable with high dilution ratios even if the original concentration is within the reported detection limit. One potential solution is to make Luria-Bertani (LB) broth directly from the sample water and inoculate it with the WCBs, but the cells may grow slower with an initial high concentration of heavy metals. Another method is to use high sample water culture ratios to minimize the difference. We tested the growth of the two biosensors in media with diluted nutrients. WCBs were grown in a medium with a nutrient concentration two times higher than that in LB broth and the same NaCl concentration to maintain electrolyte balance. Sodium chloride solution (10 g/liter) was mixed with the culture in a ratio of 5:1 or 10:1. In both cases, the growth of the two WCBs was similar to that cultured in regular LB broth. So for practical applications, the concentration detected largely represents the concentration in the original sample if using a very low dilution ratio. Due to the variation of live systems, WCBs are mainly for qualitative or semiquantitative analysis. With a low dilution ratio, the biosensors we designed can be used for an initial test of whether the sample is polluted.

Another major concern about WCBs is whether they are functional in field applications since the samples may contain highly toxic components or contaminating species, which interfere with cell growth or the accuracy of sensing. We used bacteria as the host in this work, as many other WCBs do (34–36), because bacteria do not require strict conditions for rapid growth and reproduction. They can survive at room temperature and have relatively low requirements for pH and humidity. Although relatively more robust, bacterial WCBs are still subject to environmental fluctuations since the expression of the reporter protein relies on cell growth (45, 46). To solve this issue, the biosensing components have to be decoupled from cell growth. One potential strategy is to use a cell-free expression system, and it is becoming feasible with decreasing costs. Another method is based on the differential interaction between the transcription factor and its cognate promoter in response to the analyte to develop a quick detection assay in vitro. This work has demonstrated the sensitivity of the regulatory elements in As(III) detection and the effect of positive feedback on improving the sensitivity of the biosensor. The same regulatory elements could be applied for the design of in vitro As(III) biosensors to avoid the issue of using live cells.

Overall, our results indicate that the positive feedback WCB is superior to the one lacking positive feedback in terms of response time, sensitivity, and specificity. This amplifier biosensor is able to detect arsenic below the WHO and FAO standards and could be potentially used as a test tool in situ to routinely monitor arsenic levels in drinking water. Our work shows the importance of genetic circuit engineering in improving the performance of WCBs and provides insights into the design of other biosensors. The positive feedback loop, though maybe having different effects on the gene circuit depending on the genetic context, is a useful strategy for designing and optimizing WCBs to detect other heavy metals or pollutants.

MATERIALS AND METHODS

Bacterial strains, reagents, and growth conditions.

Construction and characterization of the designed WCBs were performed in E. coli DH5α. Cells were grown in LB broth (10 g/liter peptone, 5 g/liter NaCl, 5 g/liter yeast extract). Solid plates were made using the same medium added with 1.5% (wt/vol) agar. All experiments were performed at 37°C unless otherwise noted. Antibiotics were used at the following concentrations: streptomycin (Sm) at 50 μg/ml and chloramphenicol (Cm) at 30 μg/ml.

PCR reagents, restriction endonucleases, and T4 DNA ligase were purchased from TransGen Biotech. NaAsO₂, Pb(NO₃)₂, ZnCl₂, CuCl₂, and CdCl₂ were purchased from Shandong Western Chemical Industry Co. Ltd., China. Oligo primer synthesis and sequencing were performed by Genewiz (China).

Construction of the sensor plasmids.

The plasmid of the arsenic resistance biosensor was made first by PCR amplification of the arsenite-sensing element (P_ars and arsR) (GenBank accession no. NC_000913.3) (21) using E. coli DH5α genomic DNA as the template. The primers were designed as follows with the restriction enzyme cutting sites underlined: 1F (5ʹ-CGGGATCCCTCCTTTCAAATGAATAGCC-3ʹ) and 1R (5ʹ-GGAATTCTTAACTGCAAATGTTC-3ʹ). The amplified DNA fragment was purified and digested with BamHI and EcoRI. The BamHI EcoRI DNA fragment was subsequently inserted into the pCDFDuet-1 plasmid, yielding pCDF-As. The mCherry gene was amplified from the plasmid pmCherry using primers 2F (5ʹ-GGAATTCCGTATTTAAATCAGGAGTGGAAATGGTGAAGCGGGCGAGG-3ʹ) and 2R (5ʹ-ATAAGAATGCGGCCGCCTACTTGTACAGCTCGTCCATGC-3ʹ). The resulting fragment was then cloned into the EcoRI and NotI restriction sites of the pCDF-As, yielding the plasmid pCDF-As-mCherry.

The luxR expression plasmid was constructed first by amplifying the luxR gene from the plasmid pGN68 (23) using primers 3F (5ʹ-GGAATTCAACTAAAGATTAAC-3ʹ) and 3R (5ʹ-ATAAGAATGCGGCCGCTTATTAATTTTTAAAG-3ʹ). The 304-bp DNA fragment was digested with EcoRI and NotI and then inserted into plasmid pCDF-As to give pCDF-As-luxR.

The reporter gene mCherry was amplified using primers 4F (5ʹ-GGAATTCATGGTGAGCAAGGGCGAGGAG-3ʹ) and 4R (5ʹ-CGGGATCCCTACTTGTACAGCTCGTCCATGC-3ʹ). The resulting PCR product was digested with EcoRI and BamHI and then subcloned into the respective sites of pGN68, yielding plasmid pGN68-mCherry. All constructs were confirmed by PCR/gel electrophoresis and Sanger sequencing.

Growth curve of the WCB strains.

A single colony of E. coli harboring the sensor plasmid(s) was grown overnight in LB medium containing appropriate antibiotics at 37°C. The OD₆₀₀ of the overnight cultures was adjusted to 2.0 with fresh LB medium, and then they were used as the seed to inoculate the subculture by adding 0.5 ml of the seed culture to 50 ml fresh liquid medium containing As(III) at different final concentrations (0, 0.1, 1, 10, 100, 200, 300, 400, 500, 600 µM). The subcultures were incubated at 37°C in an orbital shaker at 220 rpm. The optical density at 600 nm was first measured by spectrophotometry (UV-2000; Unico, USA) after 2 h of incubation and then measured every hour.

Fluorescence measurement.

An overnight culture with an OD₆₀₀ adjusted to 2.0 was used as the seed culture to inoculate the 5-ml subculture with a dilution rate of 1:100. The heavy metal inducer, NaAsO₂, Pb(NO₃)₂, ZnCl₂, CuCl₂, or CdCl₂, was added to a specific concentration (24). The culture was incubated at 37°C in an orbital shaker. Every hour, 200 μl of each subculture was transferred to a 96-well microplate, and the optical density at 600 nm and the fluorescence intensity were measured by the fluorescence microplate reader (M2; SpectraMax, USA) with an excitation/emission of 580/610 nm. All experiments were performed in triplicate, and the E. coli DH5α strain containing the plasmid pCDFDuet-1 was used as the negative control.

The fluorescence induction ratios (FIRs) were calculated using the formula FIR = AFU/BFU, where the arbitrary units of fluorescence (AFU) are defined as the relative fluorescence unit (RFU) divided by the optical density at 600 nm at specific arsenic concentrations and time points. The background fluorescence unit (BFU) was defined by dividing the RFU of the E. coli DH5α culture containing no metal (negative control) by its optical density at 600 nm. The fluorescence induction ratios (FIRs) of all other samples were normalized to BFU.