In recent years, directed evolution has been widely used for the development and the improvement of enzymes and proteins (Currin et al. 2015). The key to directed evolution lies in the construction of the mutant library and the efficiency of the screening process. At present, methods for constructing mutant libraries are well established. Developing efficient screening strategies has become the bottleneck of protein directed evolution technology (Cadwell and Joyce 1992; Stemmer 1994; H. Zhao et al. 1998). Traditional fluorescence activated cell sorting (FACS) screening methods have been applied to the directed evolution of transcriptional regulatory proteins, but this method requires expensive instrumentation and does not validate evolutionary results in real time (Hakkila et al. 2011). Yokobayashi et al. designed a dual genetic screening module to select a genetic inverter from a 200 fold excess of nonfunctional inverters in two rounds (Yokobayashi and Arnold 2005). The key to this design was to couple cell survival with the desired feature and maintain the evolution of the gene circuits in host cells. In this study, we designed a dual selection module to tune the specificity of a transcription factor toward a particular inducer and select the mutants that are only respond to the target inducer so that it can minimize the response to other competing inducers. In presence of the target inducer (ON selection), the mutant libraries are evolved and selected with the positive marker, the ampicillin resistance gene amp, and only the ones that can be activated by the target inducer can survive. In presence of the competing inducers (OFF selection), the library will be selected using the negative marker, the levansucrase gene sacB, and only the ones that do not respond to these inducers can survive. By alternating these ON-OFF selection steps for multiple cycles, mutant regulators only in response to the target inducer are expected to be obtained. This strategy can be applied to enhance the specificity of transcription factor based biosensors in a way that it strengthens the signal of the biosensors in response to the target analyte and minimizes the noise.
Lead is a bioaccumulative and highly toxic heavy metal that can cause serious damage to the ecological environment and human health (Bai et al. 2015). Under normal circumstances, the lead concentration in the human body should be less than 0.1 mg L− 1, and once the lead concentration exceeds the standard, it will quickly affect the nervous system and growth and development, leading to the occurrence of lead poisoning (Shen et al. 1998). Lead pollution in the environment is the main cause of frequent lead poisoning incidents (Baker et al. 1977; Oliver 1911). Therefore, the development of rapid and efficient methods to detect the lead ions concentration in the environment has become the key to the prevention and control of lead pollution. The determination of the lead concentration in the environment requires advanced chemical equipment and technical expertise, which will result in the inability to detect lead ions in real time in some areas (Badiei et al. 2013; Oliveira et al. 2011; X. Zhao et al. 2009). To solve this problem, biosensors which are simpler and less expensive than analytical instruments, and are valuable for in situ detection of lead are being used (Liao et al. 2006; Qu et al. 2016).
The use of whole-cell biosensors to determine heavy metal concentrations has been reported, and the biosensors use microbial live cells as biometric materials identify and detect substances to be tested (Aleksic et al. 2007; Tauriainen et al. 1998). At present, almost all whole-cell lead biosensors use the transcription factor PbrR from the plasmid pMOL30 of the bacterium Ralstonia metallidurans CH34 as the sensing element (Mergeay et al. 2003; Monchy et al. 2006). The lead responsive transcription factor PbrR belonging to the MerR family activates transcription upon binding to lead ions (Hobman et al. 2012). Due to the structural similarity of the MerR family transcription factors, many of them can respond to multiple divalent ions. PbrR is the most specific regulator in response to the lead ions. Still, it is known to respond to other ions like zinc, copper, mercury, etc (Angeli et al. 2004). When testing real samples, often the output signal is attributed to divalent ions as a whole, and not just lead. It is speculated that the unique physicochemical properties of the lead ions and the protein conformation of metal binding domain may be the main factors affecting the specific binding of lead ions to PbrR. As reported, the cysteine residues in regulator play a significant role in coordinating with metal ions and activating the expression of the gene that is downstream of the promoter (Monchy et al. 2006; Shewchuk et al. 1989). With limited information on the structure of PbrR, rational design of a mutant of desired function is impossible and directed evolution offers a potential solution in this situation (Bornscheuer et al. 2012; Cobb et al. 2013). However, traditional screening strategy in directed evolution does not work well to improve the specificity of these regulators. Therefore, the dual selection system was applied to improve the screening efficiency by exerting both positive and negative selection pressures.
The aim of our study was to enhance the specificity of PbrR toward lead and to mitigate the interference of the divalent metal ions zinc. To achieve this goal, we created a mutant library by error-prone PCR and evolved and selected the desired PbrR mutants with the dual selection system. Compared with the wild type, the mutant strains M1 and M2 had increased response to the lead ion with 1.8-fold and 2-fold respectively. In addition, the wild-type growth was inhibited during the OFF selection with zinc ions, while the mutant strain M1/M2 rapidly grew, and weakened zinc-binding ability was observed. Structural simulations indicated that the mutation C134R of M1 was located on the C-terminal metal-binding loop region, which may lead to the enhancement of cadmium ion binding, and the double mutations D64A and L68S of M2 were located on the α-helix α4 near the loop region of C79. Amino acid mutations near the metal binding domain of the dimeric protein may cause subtle force changes and spatial changes, leading to reduced binding capacity of zinc ions.