Figures
Abstract
The application of the luxCDABE operon of the bioluminescent bacterium Photorhabdus luminescens as a reporter has been published for bacteria, yeast and mammalian cells. We report here the optimization of fused luxAB (the bacterial luciferase heterodimeric enzyme) expression, quantum yield and its application as a reporter gene in plant protoplasts. The fused luxAB gene was mutated by error prone PCR or chemical mutagenesis and screened for enhanced luciferase activity utilizing decanal as substrate. Positive luxAB mutants with superior quantum yield were subsequently shuffled by DNase I digestion and PCR assembly for generation of recombinants with additional increases in luciferase activity in bacteria. The coding sequence of the best recombinant, called eluxAB, was then optimized further to conform to Arabidopsis (Arabidopsis thaliana) codon usage. A plant expression vector of the final, optimized eluxAB gene (opt-eluxAB) was constructed and transformed into protoplasts of Arabidopsis and maize (Zea mays). Luciferase activity was dramatically increased for opt-eluxAB compared to the original luxAB in Arabidopsis and maize cells. The opt-eluxAB driven by two copies of the 35S promoter expresses significantly higher than that driven by a single copy. These results indicate that the eluxAB gene can be used as a reporter in plant protoplasts. To our knowledge, this is the first report to engineer the bacterium Photorhabdus luminescens luciferase luxAB as a reporter by directed evolution which paved the way for further improving the luxAB reporter in the future.
Citation: Cui B, Zhang L, Song Y, Wei J, Li C, Wang T, et al. (2014) Engineering an Enhanced, Thermostable, Monomeric Bacterial Luciferase Gene As a Reporter in Plant Protoplasts. PLoS ONE 9(10): e107885. https://doi.org/10.1371/journal.pone.0107885
Editor: Hector Candela, Universidad Miguel Hernández de Elche, Spain
Received: May 6, 2014; Accepted: August 15, 2014; Published: October 1, 2014
Copyright: © 2014 Cui et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper and its supporting information files, and the nucleic acid sequence of opt-eluxAB is available from the GenBank database (accession number: KJ957766).
Funding: This work was supported by the National Natural Science Foundation of China 31170121 (to XS), 31170100 and 31370150 (to YW), and the Special fund for transgenic research by Ministry of Agriculture in China 2014ZX0800920B (to TZ). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Reporter genes are valuable tools for promoter analysis [1], imaging of gene expression [2], [3], detecting xenobiotic compounds [4], protein subcellular localization [5], protein-protein interactions [6], and discovery of genes as potential targets for disease [7]. Although the current reporter gene assay systems, such as GFP, firefly luciferase (Fluc), LacZ, CAT and GUS have greatly advanced molecular biology research, all except the bacterial luciferase system require additional manipulations such as cell lysis, substrate addition or UV excitation. A distinct advantage of the bacterial luciferase (lux) system is the ability to utilize substrates readily available in the cell (FMNH2, O2, and a fatty aldehyde), thus eliminating the need for cell lysis or exogenous substrate addition [8], [9]. Alternatively, the substrate decanal can be exogenously provided without disruption of the cell, as decanal traverses cell membranes easily. Therefore, using a bacterial luciferase-based reporter allows investigators to non-invasively study gene expression or physiological changes in viable cells in real-time. In addition, the use of a bacterial luciferase-based reporter is more convenient and economical, especially for large-scale experiments. For these reasons, development of a reliable, bacterial luciferase-based reporter for eukaryotic cells would be extremely useful especially for continuous in vivo real-time assays.
Many attempts have been made for expression of the bacterial lux system in different eukaryotic cells using either fusion proteins [10], [11] or multiple plasmids [12] in the past two decades. Unfortunately, despite its success as a bacterial reporter, development of the bacterial luciferase as a reporter for widespread use, allowing real-time detection in eukaryotic cells, has faced difficulties, such as low expression levels, lack of thermostability, and poor quantum yield. These hurdles were partially overcome through codon-optimization, addition of specialized linker regions, and co-expression of the flavin reductase gene (frp) encoding the NADPH-FMN oxidoreductase [13], but substantial light production by the lux reaction was demonstrated to occur only in the lower eukaryote Saccharomyces cerevisiae. Recently, Close et al. [14] have reported a major advance in developing an autonomous mammalian light production system by expression of all five genes (luxCDABE) of the lux operon simultaneously in a mammalian background. However, it should be noted that the bioluminescent signal from the human-optimized lux cassette was also relatively weak (several orders of magnitude lower than that of the Fluc reporter). Thus, it is clear that the bacterial luciferase could potentially benefit from further optimization to reach its full potential as a eukaryotic reporter.
In this research, to simplify the usage of the bacterial luciferase as a reporter protein in eukaryotic cells, we have constructed a new fused luxAB gene encoding a fusion protein as a reporter. Furthermore, we show that the activity of the fused bacterial luciferase was significantly increased by several steps of directed evolution, compared to the wild-type fused luxAB enzyme. The enhanced luxAB fusion gene has been successfully expressed in protoplasts of Arabidopsis and maize. These results show promise toward the potential development of a eukaryotic reporter system allowing in vivo real-time detection in the future.
Materials and Methods
Bacterial strains and vectors
Bacterial strains and plasmids used in this study are listed in Table S1. Escherichia coli and Corynebacterium glutamicum strains were cultured in Luria-Bertani (LB) broth or on LB plates at 37°C and 30°C, respectively. Yersinia pseudotuberculosis strains were cultured in Yersinia-Luria-Bertani (YLB) broth (1% tryptone, 0.5% yeast extract, 0.5% NaCl) or on YLB plates at 26°C [15]. These bacterial strains represent a wide spectrum of possible prokaryotic organisms in which a lux reporter system might be useful. When needed, antibiotics were used at the following concentrations: ampicillin, 100 µg/ml for E. coli and Y. pseudotuberculosis; chloroamphenicol, 20 µg/ml for E. coli and 10 µg/ml for C. glutamicum; kanamycin, 50 µg/ml for E. coli and Y. pseudotuberculosis, and 25 µg/ml for C. glutamicum; nalidixic acid, 30 µg/ml for C. glutamicum.
The enhanced luxAB reporter was also tested in eukaryotic cells represented by the common baker’s yeast Saccharomyces cerevisiae and in leaf protoplasts of Arabidopsis thaliana and Zea mays (maize). YPD liquid medium (1% yeast extract, 2% peptone, 2% glucose) was used for routine growth of yeast. Plasmid p425GPD used in this study contained the LEU2 selectable marker and a glyceraldehyde 3-phosphate dehydrogenase (GPD) promoter which controls the expression of exogenous proteins. S. cerevisiae strains harboring plasmids p425GPD were propagated in synthetic complete (SC) minimal media lacking leucine and containing 0.67% yeast nitrogen base (Invitrogen, Carlsbad, CA, USA) [16].
The method of protoplast preparation was optimized from previously published methodology [17]. Fully expanded leaves from 4 weeks old Arabidopsis (ecotype Columbia), growing under neutral or short photoperiods (12–13 h light or less) of 50–150 µE, were used to prepare protoplasts. Leaf strips (0.5 mm) were cut length-wise and vacuum infiltrated with digestion solution (1% cellulase, 0.2% macerozyme, 0.1% BSA, 0.035% β-mercaptoethanol, 10 mM CaCl2⋅2H2O, 0.4 M mannitol, 20 mM KCl, 20 mM MES, pH 5.7) for 30 min at −80 kPa pressure. The cells were then incubated on a rotary shaker (40 rpm) at 23°C for 1.5 h in the dark, followed by swirling at 80 rpm for 1 min. The digested cells were filtered through 75 µm mesh filters. The protoplasts were collected by centrifugation at 300 rpm for 5 min, washed once with digestion buffer, before being re-suspended with MMG solution (0.4 M mannitol; 15 mM MgCl2; 4 mM MES, pH 5.8).
For maize protoplast preparation as published [18], three day old, light-grown B73 seedings were moved to the dark at 25°C until the second leaf was about 10–15 cm long. The central 6–8 cm of the second leaf were cut into 0.5 mm strips length-wise and vacuum infiltrated with enzyme digestion solution (1.5% cellulase, 0.3% macerozyme, 0.1% BSA, 0.05% β-mercaptoethanol, 1 mM CaCl2⋅2H2O, 0.6 M mannitol, 10 mM KCl, 20 mM MES, pH 5.7) for 50 min with a pressure of −80 kPa. The digested cells were then incubated on a rotary shaker (40 rpm) at 25°C for 2 h in the dark, followed by swirling at 80 rpm for 5 min. Cells were then filtered through 45 µm mesh filters. Protoplast collection, washing, and resuspension were as described above for Arabidopsis.
Construction of plasmids
Primers used in this study are listed in Table S2. The entire coding region of the cloned pDM4-luxCDABE was used as a template to amplify a fused luxAB gene by overlap PCR generated using the gene SOEing method described by Horton et al. [19]. In the first round of PCR, the 1083 bp luxA gene and the 984-bp luxB gene PCR products were amplified using primer pairs luxA-XbaI-F/luxA-L15-R and luxB-L15-F/luxB-BglII-R. The resulting PCR products were used as template in the second round of PCR with luxA-XbaI-F and luxB-BglII-R as primers. The 2.1 Kb PCR product was recovered from an agarose gel and digested with XbaI/BglII and then ligated into XbaI/BamHI digested pBluescript II KS(+). This construct was named pBS-luxAB. The lac promoter in plasmid pUT18 and T6SS4 promoter in the enteric pathogen Y. pseudotuberculosis [15] were amplified with primer pairs pLac-SacI-F/pLac-XbaI-R and pT6SS4-SacI-F/pT6SS4-XbaI-R, respectively, and digested with SacI/XbaI and ligated into the same restriction enzyme digested pBS-luxAB and pBluescript II KS(+), resulting in plasmids pBS-pLac::luxAB, pBS-pLac, pBS-pT6SS4::luxAB and pBS-pT6SS4. To produce the expression plasmids pET28a-luxA+B and pET28a-luxAB, the bicistronic luxA+B gene amplified from pDM4-luxCDABE, and the luxAB gene fusion amplified from pBS-pLac::luxAB using the primers luxA-BamHI-F/luxB-SalI-R were cloned into compatible sites within the pET28a vector downstream of the T7 promoter, respectively.
The enhanced luxAB (eluxAB) generated by DNA shuffling described below was amplified using luxA-XbaI-F and luxB-BglII-R primers and digested with restriction enzymes XbaI/BglII and ligated into pBS-pLac digested with XbaI and BamHI. After transformation and selection on plates with ampicillin, the plasmid pBS-pLac::eluxAB was obtained. Similarly, the plasmid pXMJ19-pTac::eluxAB was constructed by inserting the eluxAB gene downstream of the tac promoter in pXMJ19 plasmid with the unique restriction sites BamHI and SalI, and the plasmid pBS-pT6SS4::eluxAB was constructed by inserting the eluxAB gene downstream of the T6SS4 promoter in pBluescript II KS(+). Plasmids pBS-pLac::eluxAB, pXMJ19-pTac::eluxAB and pBS-pT6SS4::eluxAB were electroporated into E. coli DH5α, C. glutamicum RES167 and Y. pseudotuberculosis YPIII cells, respectively.
The luxAB, eluxAB and codon optimized eluxAB (opt-eluxAB) described below were introduced into the plasmids p425GPD with the restriction sites BamHI at the 5′ end and SalI at the 3′ end to produce the plasmid p425GPD-luxAB, p425GPD-eluxAB and p425GPD-opt-eluxAB. These three plasmids were introduced into S. cerevisiae strains and selected on SC minimal selective media.
The plant expression vector was constructed as follows: For plasmid a1 construction, the eluxB fragment was generated by PCR from pBS-pLac::eluxAB using a pair of primers eluxB-NcoI-F and eluxB-R, the nos fragment was PCR-amplified using a pair of primers nos-F and nos-SalI-R from the pGL3-dual luciferase vector [20], then the eluxB-nos fragment was amplified by overlapping extension PCR using a pair of primers eluxB-NcoI-F and nos-SalI-R and using mixed eluxB and nos fragments as templates. The PCR products were digested with NcoI and SalI and cloned into the corresponding sites of the dual luciferase vector to replace the Fluc ORF. The designed primers eluxA-NheI-F and eluxA-XbaI-R, were used to PCR-amplify the eluxA fragment from pBS-pLac::eluxAB, after digested by NheI and XbaI, the fragment was cloned into the dual luciferase vector to replace another Renilla luciferase (Rluc) ORF. For plasmid a2 construction, the opt-eluxB fragment was amplified using primers opt-eluxB-NcoI-F and opt-eluxB-R, from the pUC57-Simple plasmid which includes the optimized sequence of eluxAB, the same strategy was applied to obtain the opt-eluxB-nos fragment, the primers used for amplification of opt-eluxB-nos were eluxB-NcoI-F and nos-SalI-R. The opt-eluxA fragment were amplified from pUC57-Simple plasmid by PCR-amplify using primers opt-eluxA-NheI-F and opt-eluxA-XbaI-R. The opt-eluxA fragment was cloned into a1 vector to replace the eluxA fragment. For plasmid a3 construction, the eluxAB fragment was amplified from pBS-pLac::eluxAB using primers eluxA-NheI-F and eluxB-XbaI-R. The PCR products and the a1 vector were digested by NheI and XbaI and then purified by Cycle Pure Kit (OMEGA Co., Ltd). The two purified fragments were ligated by T4 DNA ligase. For plasmid a4 construction, the opt-eluxAB fragment was obtained by digestion of pUC57-Simple plasmid by NheI and XbaI restriction enzymes and then recycled and purified. The fragment was then cloned into plasmid a3 to replace the eluxAB fragment. For plasmid a5 construction, the single copy of the CaMV35S promoter was amplified from pCAMBIA3301 using primers 35S-KpnI-F and 35S-NheI-R, the single copy of the CaMV35S promoter was used to replace the double CaMV35S promoter in a4 vector.
Overexpression and purification of recombinant proteins
Plasmids pET28a-luxAB or pET28a-luxA+B were transformed into E. coli BL21(DE3) and the recombinant bacteria containing these plasmids were grown at 37°C in LB medium to an OD600 of 0.5. The strains were then shifted to 20°C and induced with 0.4 mM isopropyl-β-D-thiogalactopyranoside (IPTG) and then cultivated for an additional 16 h at 20°C. Approximately 100 µl of culture was centrifuged, and the pellet was resuspended in 100 µl of 1×SDS sample buffer. The samples were heated at 98°C for 10 min and subjected to SDS-PAGE analysis on 12% gels made with 29∶1 acrylamide:bisacrylamide stock (Ameresco). Electrophoresis of 5 µl samples proceeded at 15 mA for about 1.5 h before staining with Coomassie blue with gentle rocking at RT for 2 h. Gels were destained in 10% acetic acid, 40% methanol at RT until the protein bands were clearly visible. For purification of 6×His-tagged fusion proteins, harvested cells were resuspended in lysis buffer (10 ml/g pellet) containing 50 mM Tris-HCl (pH 8.0), 100 mM NaCl, 0.5 mg/ml lysozyme, and 2 mM phenylmethylsulfonyl fluoride (PMSF) and then incubated on ice with occasional vortexing for 10 to 15 min until the suspension became viscous. The cells were broken by sonication on ice until the suspension was translucent. Cleared cell lysates were incubated with Ni2+ agarose beads for 2 h at 4°C, and the beads were washed with 40 times the bed volume of TBS buffer (50 mM Tris-HCl, 150 mM NaCl, pH 7.6) containing 10 mM imidazole. Recombinant 6×His-tagged LuxAB and LuxA+B proteins were eluted with TBS buffer containing 200 mM imidazole and exhaustively dialyzed against TBS to remove imidazole. The molecular weight and concentrations of the purified proteins were also analyzed using 12% SDS-PAGE gels [21].
Random mutagenesis
To increase the luciferase activity of luxAB, random mutations were introduced with either three rounds of error-prone PCR or two rounds of chemical mutagenesis. For error-prone PCR, random mutations were introduced into the amplification product during PCR by using the plasmid pBS-pLac::luxAB as template and oligos luxA-XbaI-F and luxB-BamHI-R as primers (Table S2). The mutagenesis frequency was controlled to the desired level (two to four amino acid substitutions per kb) by altering the concentration of the template and the cycle numbers. For a typical reaction, 0.2 ng template DNA were added to 100 µl of the error prone PCR system (0.2 mM of each dATP and dGTP, 1.0 mM of each dCTP and dTTP, 7 mM MgCl2, 50 mM KCl, 1 mM MnCl2, 10 mM Tris-HCl (pH 8.3), 0.3 µM of primer luxA-XbaI-F and luxB-BamHI-R, 5U Taq DNA polymerase) and program subjected to 25 PCR cycles [22], [23]. Chemical mutagenesis was performed by treating the pBS-pLac::luxAB plasmid DNA with hydroxylamine. In this instance, 5 µg of plasmid DNA was incubated in a 100 µl reaction mixture containing 0.5 M hydroxylamine, 0.5 mM Na2EDTA, and 5 mM Tris-HCl (pH 6.0) for 10 h at 37°C. The DNA was precipitated and washed thoroughly with 70% ethanol, and redissolved in ddH2O [24], [25]. The mutated luxAB produced in both error-prone PCR and chemical mutagenesis were recovered by digestion with XbaI/BamHI, and ligated into pBS-pLac digested with the same enzymes, resulting in the pBS-pLac::luxAB* mutant library and was transformed into the E. coli strain DH5α for storage and further screening.
DNA shuffling and screen assay
The substrates for the shuffling reaction were nine mutant plasmids (pBS-pLac::luxAB*) generated from the prior error prone PCR or chemical mutagenesis that showed enhanced lux quantum yield. These nine templates were each used to amplify a 2.1-kb PCR product using primers luxA-XbaI-F and luxB-BamHI-R. After the amplicons were purified, about 4 µg was digested with 0.15 unit of DNase I (Promega) in 100 µl of 50 mM Tris-HCl (pH 7.4), 1 mM MgCl2 for 10–20 min at room temperature. The digestion was loaded on a 2% low melting point agarose gel and the smear from 50 to 100 bp was purified by electrophoresis onto DE81 ion-exchange paper (Whatman) and eluted with 1 M NaCl, then precipitated by ethanol. The purified fragments were redissolved in a PCR reaction mixture at a concentration of 30 ng/µl and no primers were added at this point. Taq DNA polymerase (TaKaRa, Dalian, China) was used at 2.5 units, and ddH2O added to a total volume of 100 µl and subjected to 40–45 PCR cycles. The PCR products were purified by gel extraction. After 1∶40 dilution of these primerless PCR products into a second PCR reaction mixture, 0.3 µM of each primer luxA-XbaI-F and luxB-BamHI-R were added and 15 additional PCR cycles were applied to typically obtain a single amplicon of the expected, 2.1 kb size [26], [27].
Plasmid pBS-pLac::luxAB constructs from the random mutagenesis or subsequent DNase I recombination library, were transformed into E. coli DH5α cells and grown at 37°C to an OD600 of 0.8 in LB medium with ampicillin (100 µg/ml). 200 µl of each culture was resuspended and placed into a well of a 96-well black plate, before the rapid addition of 2 µl decanal in each well [28]. The lux activity of these assays was estimated with an Infinite M200 PRO (TECAN). Strains which had the greatest luminescence activity, relative to the original construct, were called eluxAB (enhanced luxAB) and served as templates for the next experiment.
Effects of pH and temperature on enhanced luciferase activity
The optimum pH of the best eluxAB enzyme’s activity was determined within a pH range of 4–12. The buffers were as followed: acetate (pH 4–5); phosphate (pH 6–8); glycine-NaOH (pH 9–12) [29]. The pH optimum was measured by incubating the cells in different buffers in the pH range of 4–12 at 37°C using decanal as substrate. The effect of temperature on enzyme activity was measured by incubating the enzyme from 20 to 60°C for 30 min at the previously determined pH optimum of 10. The activities of eluxAB were measured under standard assay conditions.
Complete codon optimization of eluxAB
The eluxAB sequence was codon optimized according to Arabidopsis thaliana codon usage data from the information tabulated in GenBank. The overall ratio for usage of each codon within the eluxAB gene was altered to more closely match Arabidopsis usage. The OptimumGene™ algorithm optimized a variety of parameters critical to the efficiency of gene expression, including but not limited to codon usage bias adjustment; GC content adjustment; restriction enzymes and CIS-acting elements; removal of repeat sequences and mRNA secondary structure mitigation. All efforts focused on defining a single gene sequence that would attain the highest possible level of expression in the plant cell. Once optimized, the genes were synthesized by GenScript Co., Ltd (Nanjing, China) and polyacrylamide gel (PAGE) purified to ensure full-length products. These optimized DNA fragments were ligated using TA cloning and the sequence was verified in each case [14], [30]. The codon optimized eluxAB was named opt-eluxAB and the optimized bicistronic eluxA+B was named opt-eluxA+B, respectively.
PEG mediated transformation of Arabidopsis and maize protoplasts
PEG-Ca mediated transformation followed a previously published protocol [17] with some modifications. The protoplasts of Arabidopsis were suspended in 100 µl of MMG solution at a density of 107 (and the maize protoplasts at 5×106) cells/ml in a 2 ml round-bottom centrifuge tube containing 15 µg of DNA and mixed well. An equal volume of 40% PEG solution (40% PEG4000, 0.2 M mannitol and 100 mM CaCl2⋅2H2O) was added to the protoplast-DNA mix drop-wise with gentle shaking for 25 min. The mixture was diluted with 440 µl of cell culture solution which, depending on the species, was: (Arabidopsis: 154 mM NaCl, 125 mM CaCl2⋅2H2O, 5 mM KCl, 2 mM MES pH5.7) or (maize: 0.6 M mannitol, 4 mM KCl, 4 mM MES, pH 5.7). The protoplasts were then centrifuged at 112×g, the supernatant removed, the protoplasts re-suspended in 1 ml cell culture solution, and plated in the wells of 24-well tissue culture plates (NEST Biotechnology, China). The protoplasts were incubated at 23°C (Arabidopsis) or 25°C (maize) for 18 h prior to harvest for luciferase activity determination.
Luciferase activity detection
E. coli cells transformed with pBS-pLac::luxAB or pBS-pLac::eluxAB constructs were grown at 37°C to an OD600 of 1.0 in LB broth with ampicillin (100 µg/ml). C. glutamicum cells transformed with pXMJ19-pTac::luxAB or pXMJ19-pTac::eluxAB constructs were grown at 30°C to an OD600 of 1.8 in LB broth with chloramphenicol (20 µg/ml). Y. pseudotuberculosis cells transformed with pBS-pT6SS::luxAB or pBS-pT6SS::eluxAB were grown at 26°C to an OD600 of 0.8 in YLB medium with ampicillin (100 µg/ml). Yeast cells containing plasmids p425GPD-luxAB and p425GPD-eluxAB were grown at 30°C to an OD600 of 1.0 in synthetic complete (SC) minimal medium. 100 µl of each resuspended culture was transferred into a 1.5 ml microcentrifuge tube and luminescence reactions were initiated by the addition of decanal. The activity determinations of the fused luxAB and enhanced mutants were performed on the GloMax 20/20 Luminometer (Promega) as described [31]. Protoplasts from one well of a 24 well plate were recovered by aspiration, placed in an Eppendorf tube and harvested by centrifugation at 1300 rpm for 5 min before being resuspended in 20 µl of lysis buffer (100 mM potassium phosphate, 1 mM dithiothreitol, pH 7.8). The cells were subjected to three rounds of freezing in liquid nitrogen for 30 sec, followed by thawing in a 37°C water bath for 3 min as described [14]. 200 µl detection solution (0.2 mM NAD(P)H, 80 µM FMNNa, 1% (v/v) decanal) was added, vortexed, and the solution assessed for luciferase activity. After the substrate was added, the luminescence activity was measured.
Results
Expression of the fused and non-fused luxA and luxB gene in bacteria
Cell extracts from E. coli BL21(DE3) harboring pET28a-luxAB possessed a 78-kDa luxAB fusion protein clearly present in IPTG induced bacteria when compared with the non-induced lane (Figs. 1A&B). Extracts from induced BL21(DE3) cells possessing the non-fused pET28a-luxA+B construct show a 40 kDa α subunit and a 36 kD β subunit polypeptide observed in the SDS-PAGE gel (Figs. 1A&B). The luxAB and luxA+B genes expressed under the T6SS4 promoter in Y. pseudotuberculosis cells demonstrated that the luminescence intensity of luxAB is about 0.002% that of the luxA+B (Figs. 1C&D). These data indicate a potential of improving the luxAB activity by random mutagenesis and directed evolution.
A. Schematic representation of expression constructs of luxA and luxB genes in the pET28a vector. The luxA and luxB genes either fused in one cistron or in two separate cistrons, were under the control of the T7 promoter. a1: Fusion construct of luxA and luxB genes. The ORFs of luxA and luxB were fused using a 15 amino acid linker (GGGSG)3 and the stop condon of the luxA ORF was removed. a2: the luxA and luxB ORFs were separated by an intergenic sequence and were translated independently. B. The constructs in (A) were expressed in BL21(DE3) cells. Protein expression was induced by IPTG. The cell lysate was separated by SDS-PAGE and the gel subjected to Coomassie blue staining. C. Schematic representation of expression constructs of luxA and luxB genes in the pBS-pT6SS4 vector. The constructs are similar as described in (A) except that the luxA and luxB genes were under the T6SS4 promoter. D. Comparison of luciferase activity from the LuxAB fusion protein or the separate LuxA and LuxB heterodimeric proteins in Y. pseudotuberculosis. The luminescence reaction was initiated by the addition of 1% decanal as substrate. Data are means±se, n = 3. *P<0.01.
Error-prone PCR, chemical mutagenesis and DNA shuffling
Following mutagenesis, nine luxAB mutants were picked that had been individually verified to result in higher luciferase activity than the original luxAB (Figs. 2A&B). These genes, carrying positive mutations, were selected as a pool to initiate DNA shuffling. After DNase I digestion, 50 to 100 bp fragments were recovered to ensure the best recombination frequency. The luxAB mutants were then re-assembled by PCR (Fig. 3). After two rounds of DNA shuffling, about two thousand mutant colonies were screened. One mutant was identified (eluxAB, enhanced luxAB) with remarkably greater lux activity than all others tested. Compared to the sequence of wild type luxAB, there were six mutation sites in the eluxAB gene, five of which resulted in amino acid changes (Table 1). Of these five, four were in the luxA subunit.
A. Experimental strategy for generation of luxAB mutants with increased luciferase activity. The luxAB gene was mutated by error prone PCR and chemical mutagenesis. Closed circles indicate positive mutations; open circles indicate negative mutations. B. Relative luciferase activity of the positive mutants. The luciferase activity of mutants was normalized to the wild type control. Data are means±se, n = 3. *P<0.01 vs wild type control.
Experimental strategy of DNA shuffling of luxAB of the nine positive mutants. Closed circles indicate positive mutations; open circles indicate negative mutations.
Protein modeling
The crystal structure of the Vibrio harveyi luciferase luxA and luxB holoenzyme has been resolved to 1.5 Å (1 Å = 0.1 nm) under low salt conditions [32]. Comparison of sequences showed that LuxA and LuxB from P. luminescens shared 84% and 59% amino acid identities with LuxA and LuxB from V. harveyi, respectively (Fig. S1). To investigate why the observed mutations of eluxAB enhanced luciferase activity, homology-based model structures of luxA and luxB from P. luminescens were obtained using structures of V. harveyi luciferase using the Swiss protein-modeling server [33]. Based on the model structures, none of the mutations directly affect the active site (FMN binding sites) or substrate binding residues (residues 166–233) [23]. However, the substitution of H234Y (Fig. 4) was in close proximity to the substrate-binding pockets and might indirectly affect the binding of the substrate. While the other mutation sites (Table 1) probably had no effect or only subtly changed the conformation of the active site and/or substrate specificity.
Characterization of the eluxAB
The optimal pH for eluxAB activity was distinctly centered around 10 using decanal as a substrate, with a rapid decline on either side of this value (the enzyme exhibited only 40% activity at pH 9.0 and 11.0; Fig. 5A). The temperature optimum of eluxAB at pH 10 was 40°C (Fig. 5B). Considerable activity was retained also at 30°C and 50°C after an incubation time of 30 min after which time residual activities were reduced to approximately 70% and 50%, respectively. Studies on the effect of pH and thermostability on the eluxAB clearly indicated that this enzyme is the best candidate for a wide range of applications where alkaline conditions prevail.
Use of eluxAB as a reporter in E. coli, C. glutamicum, Y. pseudotuberculosis and Yeast cells
The luxAB and eluxAB were constructed in pXMJ19-pTac, pBS-pLac, pBS-pT6SS4, and p425GPD vectors under the control of their respective promoters. The plasmids pXMJ19-pTac::luxAB and pXMJ19-pTac::eluxAB, when expressed in the Gram-positive bacterium C. glutamicum resulted in an eluxAB luminescence approximately 4-fold greater than that of luxAB (Fig. 6). Similarly, the plasmids pBS-pLac::luxAB and pBS-pLac::eluxAB were transformed into E. coli cells and the luminescence was detected. Enzyme activity of eluxAB is about five-fold higher than luxAB (Fig. 6). The greatest increase in luminescence was for Y. pseudotuberculosis cells where pBS-pT6SS4::eluxAB registered a quantum yield 40 fold greater than that of pBS-pT6SS4::luxAB (Fig. 6). Saccharomyces cerevisiae cells harboring p425GPD-luxAB or p425GPD-eluxAB had an eluxAB activity approaching 10-fold that of luxAB (Figs. 7A&B). Thus we conclude that eluxAB can be used as a robust reporter in a variety of prokaryotic and simple eukaryotic species.
Data are means±se, n = 3. *P<0.01 vs wild type control.
A. Schematic representation of expression vectors for luxAB, eluxAB, and opt-eluxAB gene fusions in yeast cells. B. Quantum yield of LuxAB, eLuxAB and opt-eLuxAB in yeast.
Codon optimization of the eluxAB gene for Arabidopsis expression and use of opt-eluxAB as a reporter in Arabidopsis and maize protoplasts
In order to maximize the expression of eluxAB in Arabidopsis and maize protoplasts, the codons of eluxAB were optimized according to Arabidopsis codon bias (Fig. S2). The optimized eluxAB gene opt-eluxAB (Genbank accession number: KJ957766) was also transformed into S. cerevisiae cells in vector p425GPD. Surprisingly, compared to the luciferase activity of eluxAB and luxAB in yeast, the quantum yield of opt-eluxAB was least (Fig. 7B). The possible reason was that Arabidopsis codon biased opt-eluxAB wasn’t suitable for expression in yeast cells. The monocistronic eluxAB, opt-eluxAB and the bicistronic eluxA+B, opt-eluxA+B were constructed in the pGL3 basic vector under the control of two copies of the CaVM35S promoter (Fig. 8A). The monocistronic opt-eluxAB cassette was also placed under the control of a single copy of the CaMV35S promoter. The plant expression vectors were transformed into Arabidopsis and maize protoplasts using PEG and, following incubation, the luminescence was detected. Compared to the non-codon-optimized eluxA+B or eluxAB genes, the opt-eluxA+B or opt-eluxAB showed dramatically increased luminescence activity in Arabidopsis cells (Fig. 8B). When these mono- or bi-cistronic eluxAB or opt-eluxAB vectors were tested in maize protoplasts, opt-eluxAB luminescence was approximately 3 fold greater than that of the other three constructs (Fig. 8C) although it was 6 times less than the same construct in Arabidopsis protoplasts (compare Figs. 8B&C) for which it was designed. To further determine if opt-eluxAB could be applied as a reporter gene in Arabidopsis cells, the opt-eluxAB gene was driven by one copy or two copies of 35S promoter and transformed into Arabidopsis protoplasts. The luminescence of opt-eluxAB under the regulation of two copies of 35S promoter was significantly greater than that produced when under the control of one copy of the 35S promoter (Fig. 8D).
A. Schematic representation of expression vectors for eluxAB, or opt-eluxAB gene fusions in plant protoplasts driven by one or two copies of the CaMV35S promoter. B. Comparison of luciferase activity of eLuxA+B, eLuxAB, opt-eLuxA+B and opt-eLuxAB in Arabidopsis. C. Comparison of luciferase activity of eLuxA+B, eLuxAB, opt-eLuxA+B and opt-eLuxAB in maize. D. Comparison of opt-eLuxAB luminescence under different promoter strengths.
To compare the luminescence activity of opt-eluxAB with Rluc or Fluc reporters, luminescence activities were compared in Arabidopsis and maize protoplast cells transformed with Rluc and Fluc reporter vector (pGL3-dual luciferase) or opt-eluxAB reporter vector (pGL3-opt-eluxAB). The luminescence activity of opt-eluxAB is comparable to Rluc, though it is significantly lower than Fluc (Table S3). This result indicates that the bacterial luciferase has the potential to be applied as a new reporter in plant cells. Since the toxicity of decanal to Arabidopsis protoplast cells was negligible at the detection concentration (1%, Fig. S3), the opt-eluxAB reporter can be used for real-time detection in viable plant protoplast cells.
Discussion
Engineering of an enhanced, monomeric bacterial luciferase eluxAB gene from P. luminescens as a reporter in prokaryotes and eukaryotes provides researchers a unique tool allowing real-time monitoring of exogenous gene expression from whole cells without lysis. A fused luxAB gene from Photorhabdus luminescens has been cloned and successfully expressed in yeast [16] and mammalian cells [14], [30] previously, but the luminescence intensity was limited, despite exogenous substrate addition or codon-optimization [9], [16]. In order to overcome these problems and develop the bacterial luciferase as a convenient reporter, here we demonstrate for the first time that the fused luxAB gene could be directed evolved using a series of random mutations and shuffling techniques to produce an eluxAB with increased luciferase activity in prokaryotic cells. Testing for enhanced luxAB activity in prokaryotic systems permitted rapid evaluation of the efficacy of thousands of mutants. Subsequently, the best of these mutants was successfully expressed as a codon-optimized form of the eluxAB gene in Arabidopsis and maize protoplasts. The eluxAB gene produced by shuffling has several advantages. First, the fused eluxAB derived from P. luminescens is easy to construct and convenient to detect due to the facile application of exogenous substrate (decanal), which easily permeates into the cells [8] and does not require cell lysis for eluxAB detection in prokaryotes and yeast cells [11]. Upon addition of the aldehyde substrate, the activity of eluxAB can be followed in vivo by measuring light emission [14], [31]. In contrast, bioreporters constructed using lacZ, cat, or Fluc, for example, require a succession of pre-processing steps involving cell lysis, permeabilization and/or the addition of secondary substrates before final signal generation [34], [35]. Second, compared to the reporter GFP, eluxAB could be measured without the need to generate, and yet exclude, excitation wavelengths, avoiding some problems such as photonic bleaching and high background [36], [37]. Third, previous work with lux genes isolated from P. luminescens has demonstrated that the luciferase is thermostable at temperatures as high as 45°C [38], which is a greater thermal stability than that of the V. harveyi or V. fischeri luciferase enzymes. Therefore, the bacterial luciferase system seems more suitable to the study of environmental or developmental changes in gene expression. High activity at alkaline pH is normally regarded as an appreciable character for the industrial production [29].
Directed evolution has become a powerful strategy for improving the activity of enzymes in a targeted manner coupling various methods such as error-prone PCR and chemical mutagenesis to generate large variant libraries [39]. In this study random mutagenesis and DNA shuffling of luxAB provides a directed approach to improve the activity of the fused bacterial luciferase. As a result, the eluxAB strain was produced, which shows considerably improved luciferase activity in a variety of organisms. There are five mutations resulting in amino acid substitutions and one silent mutation that occurred to result in eluxAB. Previous work on the luciferase enzyme determined that the lux catalytic properties were primarily determined by the luxA subunit [40]. In this paper, of the five missense mutations, four appeared in luxA. None of the mutations occurs in the catalytic active site or appears to directly affect substrate binding. Because the amino acid sequences of luxA and luxB are highly conserved according to the result of alignments of amino acid sequences of luxA and luxB from P. luminescens and V. harveyi (Fig. S1), it is likely that if mutations occurred in the active site, they would inactivate the luciferase. Therefore, although it is speculative, it is most probable that the mechanism resulting in enhanced lux activity involves mutations that exert their effects by indirectly changing the conformation of the aldehyde binding sites rather than by interacting with substrates directly.
The bacterial luciferase eluxAB can be assayed with high sensitivity compared to the luxAB both in intact bacterial- and yeast-cells. In addition, the opt-eluxAB or opt-eluxA+B has a greater luciferase activity than the eluxAB and eluxA+B in Arabidopsis cells. Evidently, codon optimization of eluxAB led to an increased luminescence. Greater luminescence may be caused by greater lux expression levels, greater luciferase activity due to superior folding of the tethered luxA and luxB subunits, higher quantum yield per catalytic round, or any of a plethora of different reasons or combinations thereof. The luciferase activity of monocistronic opt-eluxAB is greater than that of opt-eluxA and opt-eluxB expressed in two separate cistrons when expressed from the same promoter (Figs. 8B&C). It is tempting to speculate that the separate subunits cannot be assembled efficiently into a functional holoenzyme in Arabidopsis cells. What’s more, under the regulation of the 2×35S promoter, the luminescence of opt-eluxAB is about two times greater than that from one copy of the 35S promoter. This is compelling evidence that the opt-eluxAB can eventually be developed as a useful marker gene for the quantitative assessment and detection of different activities from promoters in transgenic plants.
The eluxAB is detectable using currently available technologies and offers prolonged expression without cell lysis in both prokaryotes and yeast [11]. The lux system derived from bacteria is particularly useful as a prokaryotic reporter because the luxC, luxD, and luxE genes can provide continuous supplies of the aldehyde substrate of lux within the cells without any external manipulations [14]. Of course, endogenous luxA and luxB genes within this operon may provide background lux readings precluding its use in these systems. However, modifications of the reporter system to include the other genes within the bacterial lux operon in the plant of interest may produce greater lux activity.
The plant-adapted bacterial luciferase gene opt-eluxAB also has some disadvantages. For example plant cells do not contain sufficient FMNH2 to drive the opt-eluxAB catalyzed reaction at Vmax, because FMN is enzyme-bound or enclosed in cell compartments in plant cells [34]. However, this is not a serious concern because, upon exogenous application of FMNH2, plant cells expressing luxAB may produce greater quantum yields [34], [41]. However, our ultimate goal is to have efficient expression and activity without the need of cell lysis or exogenous substrate application. One possible means to circumvent the poor availability of FMNH2 in the cytoplasm will be to target the opt-eluxAB into an organelle containing this cofactor using a target peptide on the N-terminus of opt-eluxAB. Further studies are needed to determine the best means by which to express the frp gene in vivo to produce enough FMNH2 in plant cells. Additionally, the basic pH optimum (10) of the eluxAB enzyme, and the abrupt attenuation of eluxAB activity to either side of this optimum is a major hurdle for the current use of this system in eukaryotic cells. Reducing the pH optimum of the eluxAB enzyme closer to that of the neutral pH of the cell is the current major focus of our synthetic biology approach using similar techniques to those described here. Alterations of pH optima and/or a broadening of the pH optimum, of several pH units, using directed evolution has been possible for a number of commercial applications [42], [43].
In a word, the eluxAB and opt-eluxAB demonstrate that, using directed evolution, synthetic biology, and by paying attention to codon bias, we can greatly increase the luminescence of a bacterial luciferase in plant cells, thus making it more suitable and convenient to use in these eukaryotes. This demonstration provides the foundation for further alterations in the opt-eluxAB gene to produce a reporter suitable as a practical bioprocess monitor, as a high-throughput promoter expression screening technique and in applications for medical diagnosis.
Supporting Information
Figure S1.
Pairwise sequence alignment of the amino acid sequences between luxA and luxB genes from P. luminescens and V. harveyi.
https://doi.org/10.1371/journal.pone.0107885.s001
(TIF)
Figure S2.
Alignment of eluxAB gene (Un-optimized) and codon-optimized eluxAB gene (Optimized) sequences. Base changes are indicated in orange. The six mutation sites in the eluxAB gene are indicated with the black hollow box.
https://doi.org/10.1371/journal.pone.0107885.s002
(TIF)
Figure S3.
The tolerance of Arabidopsis protoplasts to decanal treatment. Protoplasts treated with different concentrations of decanal were photographed under optical microscope at time 0, 1 and 4 hours, respectively. Row 5 shows the protoplast cells tested with 200 µg/ml kanamycin as a control. Kanamycin but not decanal treatment leads to quick lysis of protoplast cells.
https://doi.org/10.1371/journal.pone.0107885.s003
(TIF)
Table S1.
Bacterial strains and plasmids used in this study.
https://doi.org/10.1371/journal.pone.0107885.s004
(DOCX)
Table S3.
Comparison of luminescence activities between bacterial, Renilla and firefly luciferases in plant protoplasts. Luminescence activities were measured in Arabidopsis and maize protoplast cells transformed with Rluc and Fluc reporter vector (pGL3-dual luciferase) or with opt-eluxAB reporter vector (pGL3-opt-eluxAB). The number of Arabidopsis protoplasts used was about 5×106 and of the maize protoplasts was about 3×105 in 100 µl volume culture.
https://doi.org/10.1371/journal.pone.0107885.s006
(DOCX)
Acknowledgments
We thank Dr. Bruce Downie (University of Kentucky) for critical reading of the manuscript and for helpful comments. We are grateful to Prof. Shuang-Jiang Liu at Institute of Microbiology, Chinese Academy of Sciences, Dr. Zhao-Qing Luo at Purdue University, and Prof. Paul Williams, Prof. Miguel Cámara and Dr. Steve Atkinson at University of Nottingham, for their generous gift of strains or plasmids.
Author Contributions
Conceived and designed the experiments: XS TZ YW. Performed the experiments: BC LZ YS JW CL TW. Analyzed the data: XS TZ BC LZ. Contributed reagents/materials/analysis tools: XS TZ YW. Contributed to the writing of the manuscript: XS TZ YW BC.
References
- 1. Xiong AS, Peng RH, Zhuang J, Davies J, Zhang J, et al. (2012) Advances in directed molecular evolution of reporter genes. Crit Rev Biotechnol 32: 133–142.
- 2. Kadurugamuwa JL, Sin LV, Yu J, Francis KP, Kimura R, et al. (2003) Rapid direct method for monitoring antibiotics in a mouse model of bacterial biofilm infection. Antimicrob Agents Chemother 47: 3130–3137.
- 3. Kuklin NA, Pancari GD, Tobery TW, Cope L, Jackson J, et al. (2003) Real-time monitoring of bacterial infection in vivo: development of bioluminescent staphylococcal foreign-body and deep-thigh-wound mouse infection models. Antimicrob Agents Chemother 47: 2740–2748.
- 4. DiGrazia PM, King JM, Blackburn JW, Applegate BA, Bienkowski PR, et al. (1991) Dynamic response of naphthalene biodegradation in a continuous flow soil slurry reactor. Biodegradation 2: 81–91.
- 5. Monshausen GB (2012) Visualizing Ca2+ signatures in plants. Curr Opin Plant Biol 15: 677–682.
- 6. Kerppola TK (2006) Design and implementation of bimolecular fluorescence complementation (BiFC) assays for the visualization of protein interactions in living cells. Nat Protocols 1: 1278–1286.
- 7. Taghva A, Kim PE, Liu CY, Apuzzo ML (2010) Molecular imaging, part 1: apertures into the landscape of genomic medicine. World Neurosurgery 73: 307–316.
- 8. Meighen EA (1991) Molecular biology of bacterial bioluminescence. Microbiol Rev 55: 123–142.
- 9. Close D, Xu T, Smartt A, Rogers A, Crossley R, et al. (2012) The evolution of the bacterial luciferase gene cassette (lux) as a real-time bioreporter. Sensors 12: 732–752.
- 10. Almashanu S, Musafia B, Hadar R, Suissa M, Kuhn J (1990) Fusion of LuxA and LuxB and its expression in E. coli, S. cerevisiae and D. melanogaster. J Biolumin Chemilumin 5: 89–97.
- 11. Boylan M, Pelletier J, Meighen E (1989) Fused bacterial luciferase subunits catalyze light emission in eukaryotes and prokaryotes. J Biol Chem 264: 1915–1918.
- 12. Koncz C, Olsson O, Langridge WH, Schell J, Szalay AA (1987) Expression and assembly of functional bacterial luciferase in plants. Proc Natl Acad Sci U S A 84: 131–135.
- 13. Lei B, Tu S-C (1998) Mechanism of reduced flavin transfer from Vibrio harveyi NADPH-FMN oxidoreductase to luciferase. Biochemistry 37: 14623–14629.
- 14. Close DM, Patterson SS, Ripp S, Baek SJ, Sanseverino J, et al. (2010) Autonomous bioluminescent expression of the bacterial luciferase gene cassette (lux) in a mammalian cell line. PLoS One 5: e12441.
- 15. Zhang W, Wang Y, Song Y, Wang T, Xu S, et al. (2013) A type VI secretion system regulated by OmpR in Yersinia pseudotuberculosis functions to maintain intracellular pH homeostasis. Environ Microbiol 15: 557–569.
- 16. Gupta RK, Patterson SS, Ripp S, Simpson ML, Sayler GS (2003) Expression of the Photorhabdus luminescens lux genes (luxA, B, C, D, and E) in Saccharomyces cerevisiae. FEMS Yeast Res 4: 305–313.
- 17. Yoo SD, Cho YH, Sheen J (2007) Arabidopsis mesophyll protoplasts: a versatile cell system for transient gene expression analysis. Nat Protocols 2: 1565–1572.
- 18. Sheen J (1991) Molecular mechanisms underlying the differential expression of maize pyruvate, orthophosphate dikinase genes. Plant Cell 3: 225–245.
- 19. Horton RM, Hunt HD, Ho SN, Pullen JK, Pease LR (1989) Engineering hybrid genes without the use of restriction enzymes: gene splicing by overlap extension. Gene 77: 61–68.
- 20. Gu L, Han Z, Zhang L, Downie B, Zhao T (2013) Functional analysis of the 5′ regulatory region of the maize galactinol synthase2 gene. Plant Sci 213: 38–45.
- 21. Xu S, Peng Z, Cui B, Wang T, Song Y, et al. (2013) FliS modulates FlgM activity by acting as a non-canonical chaperone to control late flagellar gene expression, motility and biofilm formation in Yersinia pseudotuberculosis.. Environ Microbiol 16: 1090–1104.
- 22.
Cirino PC, Mayer KM, Umeno D (2003) Generating mutant libraries using error-prone PCR. Directed Evolution Library Creation. Springer. 3–9.
- 23. Hosseinkhani S, Szittner R, Meighen E (2005) Random mutagenesis of bacterial luciferase: critical role of Glu175 in the control of luminescence decay. Biochem J 385: 575–580.
- 24. Rasila TS, Pajunen MI, Savilahti H (2009) Critical evaluation of random mutagenesis by error-prone polymerase chain reaction protocols, Escherichia coli mutator strain, and hydroxylamine treatment. Anal Biochem 388: 71–80.
- 25. Kaur J, Sharma R (2006) Directed evolution: an approach to engineer enzymes. Crit Rev Biotechnol 26: 165–199.
- 26. Stemmer WP (1994) DNA shuffling by random fragmentation and reassembly: in vitro recombination for molecular evolution. Proc Natl Acad Sci U S A 91: 10747–10751.
- 27. Zuo Z-Y, Zheng Z-L, Liu Z-G, Yi Q-M, Zou G-L (2007) Cloning, DNA shuffling and expression of serine hydroxymethyltransferase gene from Escherichia coli strain AB90054. Enzyme Microb Technol 40: 569–577.
- 28. Escher A, O’Kane DJ, Lee J, Szalay AA (1989) Bacterial luciferase alpha beta fusion protein is fully active as a monomer and highly sensitive in vivo to elevated temperature. Proc Natl Acad Sci U S A 86: 6528–6532.
- 29. Niladevi KN, Jacob N, Prema P (2008) Evidence for a halotolerant-alkaline laccase in Streptomyces psammoticus: Purification and characterization. Process Biochemistry 43: 654–660.
- 30. Patterson SS, Dionisi HM, Gupta RK, Sayler GS (2005) Codon optimization of bacterial luciferase (lux) for expression in mammalian cells. J Ind Microbiol Biotechnol 32: 115–123.
- 31. Tinikul R, Thotsaporn K, Thaveekarn W, Jitrapakdee S, Chaiyen P (2012) The fusion Vibrio campbellii luciferase as a eukaryotic gene reporter. J Bacteriol 162: 346–353.
- 32. Fisher AJ, Thompson TB, Thoden JB, Baldwin TO, Rayment I (1996) The 1.5-Å resolution crystal structure of bacterial luciferase in low salt conditions. J Biol Chem 271: 21956–21968.
- 33. Schwede T, Kopp J, Guex N, Peitsch MC (2003) SWISS-MODEL: an automated protein homology-modeling server. Nucleic Acids Res 31: 3381–3385.
- 34. Koncz C, Langridge WH, Olsson O, Schell J, Szalay AA (1990) Bacterial and firefly luciferase genes in transgenic plants: advantages and disadvantages of a reporter gene. Developmental Genetics 11: 224–232.
- 35. Miller J (1993) A short course in bacterial genetics: a laboratory manual and handbook for Escherichia coli and related bacteria. Trends Biochem Sci 18: 193.
- 36. Dickson RM, Cubitt AB, Tsien RY, Moerner W (1997) On/off blinking and switching behaviour of single molecules of green fluorescent protein. Nature 388: 355–358.
- 37. Cubitt AB, Heim R, Adams SR, Boyd AE, Gross LA, et al. (1995) Understanding, improving and using green fluorescent proteins. Trends Biochem Sci 20: 448–455.
- 38. Szittner R, Meighen E (1990) Nucleotide sequence, expression, and properties of luciferase coded by lux genes from a terrestrial bacterium. J Biol Chem 265: 16581–16587.
- 39. Turner NJ (2003) Directed evolution of enzymes for applied biocatalysis. Trends Biotechnol 21: 474–478.
- 40. Valkova N, Szittner R, Meighen EA (1999) Control of luminescence decay and flavin binding by the LuxA carboxyl-terminal regions in chimeric bacterial luciferases. Biochemistry 38: 13820–13828.
- 41. Kirchner G, Roberts J, Gustafson G, Ingolia T (1989) Active bacterial luciferase from a fused gene: expression of a Vibrio harveyi luxAB translational fusion in bacteria, yeast and plant cells. Gene 81: 349–354.
- 42. Bessler C, Schmitt J, Maurer KH, Schmid RD (2003) Directed evolution of a bacterial α-amylase: Toward enhanced pH-performance and higher specific activity. Protein Sci 12: 2141–2149.
- 43. Chen KC, Wu CH, Chang CY, Lu WC, Tseng Q, et al. (2008) Directed evolution of a lysosomal enzyme with enhanced activity at neutral pH by mammalian cell-surface display. Chem Biol 15: 1277–1286.