A Thermostable β-Glucuronidase Obtained by Directed Evolution as a Reporter Gene in Transgenic Plants

A β-glucuronidase variant, GUS-TR3337, that was obtained by directed evolution exhibited higher thermostability than the wild-type enzyme, GUS-WT. In this study, the utility of GUS-TR337 as an improved reporter was evaluated. The corresponding gus-tr3337 and gus-wt genes were independently cloned in a plant expression vector and introduced into Arabidopsis thaliana. With 4-MUG as a substrate, plants containing the gus-wt gene showed no detectable β-glucuronidase activity after exposure to 60°C for 10 min, while those hosting the gus-tr3337 gene retained 70% or 50% activity after exposure to 80°C for 10 min or 30 min, respectively. Similarly, in vivo β-glucuronidase activity could be demonstrated by using X-GLUC as a substrate in transgenic Arabidopsis plants hosting the gus-tr3337 gene that were exposed to 80°C for up to 30 min. Thus, the thermostability of GUS-TR3337 can be exploited to distinguish between endogenous and transgenic β-glucuronidase activity, which is a welcome improvement in its use as a reporter.


Introduction
There are about fifty reporter genes that have proved useful in transgenic plant research, based on their efficiency, scientific applications and commercialization. Reporter genes serve as indicators to study transgenic events, for example by facilitating visual identification of successfully transformed entities in a large background of non-transformed material. Although many reporter genes have been proposed, only a few of them have been commercialized [1]. b-glucuronidase (GUS) [2,3], green fluorescent protein (GFP) [4,5], luciferase (LUC) [6,7,8] are three important reporters that have a proven track record in transgenic plant research.
The advantage of GFP and LUC as reporters includes the possibility of monitoring in live tissues and in real time [1,4,8]. These two reporters allow continuous monitoring of the gene activity even through the developmental stages of the transgenic plants [1,5,9]. However, there are several disadvantages in the use of these two reporter genes. For example, the half life of LUC is short (a few hours) in plants and the process of catalyzing the release of oxyluciferin is slow [6,7,10]. In case of GFP, there may be interference due to a high background of auto-fluorescence. Another major drawback in the use GFP as a reporter is that its detection requires relatively expensive instrumentation, namely the fluorescence microscope.
The bacterial enzyme GUS encoded by the E. coli gusA gene is to date the most widely used as a reliable transgenic reporter system in transgenic plant research [1]. GUS has some advantages as a reporter in transgenic plant research and commercialization. First, in genetically modified plants E.coli GUS can be regarded as safe for environment and consumers [11]. Second, in transgenic plants, overexpressed GUS is very stable and shows no toxicity in plant growth and development. In addition, detection of GUS does not require specialized equipment as it can easily be observed after staining with substrate such as 5-bromo-4-chloro-3-indolyl-b-Dglucuronic acid (X-GLUC). Thus, the GUS reporter system is a simple, reliable and cost-effective method to be used in transgenic studies [1,12]. Despite these benefits of the gusA gene, falsepositives are difficult to eliminate in some plants due to endogenous GUS activity [13]. Increasing the thermal stability of GUS through directed evolution in vitro could potentially eliminate false-positives and dramatically increase its veracity as a reporter.
Directed evolution in vitro via DNA shuffling, is an important method for generating proteins with improved functions. Stemmer introduced the method of in vitro DNA shuffling for the formation of recombinant genes from a set of parental genes [30,31]. We have previously reported the directed evolution of a GUS variant, GUS-TR3337, which is significantly more resistant to high temperatures than wild-type enzyme, GUS-WT [32,33]. Here, our objective was to test the utility of GUS-TR3337 as an improved reporter gene in a transgenic plant.
In order to study the activity of the thermostable GUS in plants, it was introduced into Arabidopsis thaliana by the floral dip method. The transgenic Arabidopsis plants containing the wild-type gus-wt gene (Genbank No. AF485783), lost most of the GUS activity within 5 min of exposure to 60uC. In contrast, Arabidopsis plants transformed with the gus-tr3337 gene (GenBank No. DQ513152), retained the activity when heated upto 80uC for as long as 30 min. We confirm here that the mutant thermostable GUS gene, gus-tr3337 is an improved reporter for transgenic research.

Materials and Methods
Chemicals, enzymes, plasmids and yeast strains All reagents were purchased from Sigma Chemical Co. Ltd (St Louis, MO, USA) and Sinopharm Chemical Reagent Co. Ltd (Shanghai, China). Restriction enzymes were purchased from Promega (Madison, USA). High-fidelity DNA polymerase Pyrobest TM was purchased from Takara Co., Ltd (Dalian, China). DNA sequencing kit was purchased from Applied Systems Company (Foster City, CA, USA). QIAquick TM gel extraction kit was purchased from Qiagen (Stanford, CA, USA). X-GLUC and plasmid pBI121, containing the gusA gene, were purchased from Clontech (Palo Alto, CA, USA). E. coli strain DH5a used for amplification of plasmid DNA was purchased from Stratagene (La Jolla, CA, USA).

Construction of plant expression vectors and plant transformation
The original gus-wt and mutant gus-tr3337 genes were independently cloned in the binary vector pYF7716 that was constructed by our research team. This vector contains doublecauliflower mosaic virus 35S promoter, to direct transcription of the transgene as well as that of the intron-containing neomycin phosphotransferase II gene (Km). In addition, the vector carries the omega enhancer and the nopaline synthase terminator, that were fused to the tobacco SAR (scaffold attachment region) ( Figure S1). The gus-wt and gus-tr3337 constructs generated in this vector were introduced into Agrobacterium tumefaciens GV3101 by electroporation. A. thaliana cv. Columbia was transformed by the floral dip method as described previously [34]. The presence of the gus transgenes was confirmed via PCR amplification of the genes in Arabidopsis plants using the following specific primers pairs: GUS-WTF1 (59-ATGTTACGTCCTGTAGAAACCCCAACC-39) and GUS-WTR1 (59-TCATTGTTTGCCTCCCTGCTGCGGTTT-TTC-39), for the gus-wt gene, and GUS-TRF1 (59-TCACGCCG-TACG TTATTGCCGGG-39) and GUS-TRR1 (59-TTGCT-ACCGCCAACGCGTAATATG-39) for the gus-tr3337 gene. The PCR conditions were: denaturation for 10 min at 94uC, followed by 30 cycles, each of 40 s at 94uC, 30 s at 58uC, 2 min at 72uC; and then extension for 10 min at 72uC. Standard DNA manipulation methods were used in plasmid construction and E. coli transformation [35].

Histochemical analysis of GUS activity in transgenic Arabidopsis plants
For the analysis of expression of the gus transgenes, seeds of the trangenic Arabidopsis plants were sterilized with 1% (v/v) NaOCl solution and sown in pots containing a 1:1:1 mixture of peat moss/perlite/vermiculite. Pots were kept in a controlled-environment growth chamber at 22uC, programmed for a photope-riod of 16 h-day under cool white fluorescent light and 8 h-dark. High temperature treatments were performed on whole two-week old seedlings by exposure to temperatures of 70uC, 80uC, 30uC for different times (10 min, 20 min, 30, min). The heat treatment was given by placing a single seedling in a 0.6 mL capacity Eppendorf TM tube containing 150 mL ddH 2 0, and then holding the tube at the desired heating regime in a PCR System PTC-200, using the lid heating function to avoid the water evaporation. Then, the water was removed. Based on the apparent optimal temperature of b-glucuronidase activity tested in transgenic Arabidopsis plants, the plant samples were incubated at 45uC for transgenic gus-wt Arabidopsis samples or 60uC for transgenic gus-tr3337 Arabidopsis samples, with 1 mg/mL X-GLUC in a GUS activity buffer (50 mM sodium phosphate, pH 7.0; 1 mM EDTA; 0.1% Triton X-100; 0.5 mM K3[Fe(CN)6]; 0.5 mM K4[Fe(CN)6]?3H2O). Histochemical staining for GUS activity was performed essentially as described by Jefferson et al. [2] and Sudan et al. [36] for most experiments. The stained samples were immersed in 70% ethanol to remove chlorophyll. The ethanol solution was changed three times at 1-h intervals. Extracts of whole three-week-old seedlings in GUS extraction buffer were used for the analysis of enzyme kinetics. Seedlings were frozen by immersion in liquid nitrogen and ground with a pestle and mortar in the presence of liquid nitrogen. GUS extraction buffer, 500 mL, was added and the plant material was vigorously ground to a finely pulverized powder. The extract thus obtained was centrifuged at 8,0006 g for 1 min at 4uC. The supernatant was recovered and diluted one-hundred fold with GUS extraction buffer in order to decrease the effect of any potential indigenous GUS inhibitors [39]. The diluted extract, 10 mL was added to 4-MUG, 90 mL (2 mM), to initiate the reaction. After 1 h the reaction was stopped by the addition of 900 mL of 0.2 M Na 2 CO 3 , and fluorescence was measured using a DyNA Quant 200 TM spectrofluorometer.

GUS enzyme kinetics by fluorometric assay in transgenic
The optimal pH for the activity of enzymes GUS-WT and GUS-TR3337 from transgenic Arabidopsis plants was determined at 37uC at every half point between pH values 2.5 and 13.0. The activity of both enzymes from transgenic Arabidopsis plants was measured at 5-degree intervals between 25uC and 90uC to determine the optimum temperature at which maximum activity was retained. The thermostability of the two enzymes from transgenic Arabidopsis plants was determined by measuring the residual activity at 37uC after subjecting the samples to temperatures of 60uC, 70uC, 80uC and 90uC for different times (10 min, 20 min and 30 min), followed by chilling on ice. All reactions were performed in triplicate. Kinetic parameters were determined as described in previous reports [32,40,41].

Over-expression of gus-wt and gus-tr3337 in Arabidopsis
The cloned gus-wt or gus-tr3337 genes under the control of the CaMV 35S promoter were independently introduced into Arabidopsis ( Figure S1). The transgenic Arabidopsis plants thus recovered showed high levels of expression of the two genes ( Figure S2). There were no noticeable differences in the morphology or growth among the above transgenic plants and the parental Arabidopsis plants (data not shown).

Thermostability of the mutant GUS-TR3337 in transgenic Arabidopsis plant
High temperature treatments were performed with two-weekold whole seedlings to evaluate the thermostability of the mutant GUS-TR3337 in transgenic Arabidopsis plants. Seedlings of wildtype Arabidopsis could not been stained by X-GLUC when they were not or subjected to 60uC for 5 min (Figure 1). Seedlings containing the wild-type gus-wt gene or the mutant gus-tr3337 gene exhibited stable GUS activity when they were not subjected to high temperature treatments (Figures 2A and 3A). The seedlings expressing the gus-wt gene were stained weakly by X-GLUC after being subjected to 60uC for 5 min or 10 min (Figures 2B and 2C) and the enzyme appeared to be inactivated by treatments for longer duration or at higher temperature ( Figures 2D-F).
On the opposite, seedlings expressing the mutant gus-tr3337 gene exhibited stable GUS activity when treated at 60uC and 70uC for 10 min, 20 min or 30 min ( Figures 3C-E and F-H), and even at 80uC for 10 min, 20 min or 30 min (Figures 3I-K). However, treatment at 90uC for 10 min or longer appeared to inactivate the GUS-TR3337 enzyme (Figures 3H-I).

Characterization of the GUS-TR3337 activity in transgenic Arabidopsis plants
Three-week-old transgenic Arabidopsis seedlings containing the mutant gus-tr3337 gene or the wild-type gus-wt gene were used to characterize the GUS-TR3337 and GUS-WT enzymes in seedling extracts. Three representative seedlings each (P1, P2 and P3) expressing either of transgenes were tested using 4-MUG as substrate at 37uC. A single activity peak was found at pH 6.5 in all six transgenic Arabidopsis lines expressing gus-tr3337 or gus-wt gene ( Figure 4A).
The apparent optimal temperature of GUS activity tested at pH 7.0 in extracts of transgenic Arabidopsis plants over-expressing gus-wt gene was 45uC, while that of transgenic Arabidopsis plants over-expressing gus-tr3337 gene was 60uC ( Figure 4B). The GUS-WT activity in transgenic Arabidopsis lines at 45uC was about 30% higher than that at 35uC. The GUS-TR3337 activity in transgenic Arabidopsis lines at 60uC was .2 times of its activity at 35uC.
The GUS-WT activity in the extracts of transgenic Arabidopsis decreased significantly after being subjected to 60uC for 10 min, and was undetectable after treatment at 70uC or higher temperatures for 10 min or longer (Figure. 5). In contrast, GUS-TR3337 activity in the transgenic Arabidopsis plants was much more resistant to heat treatment. Nearly 90% and 70% of GUS activity was retained when plants were heated at 70uC for 10 min or 30 min, and about 70% and 50% was retained when heated at 80uC for 10 min or 30 min.
Similarly, when heated at 70uC for 10 min or 30 min, they retained nearly 90% and 70% of this activity, respectively; and when heated at 80uC for 10 min or 30 min, retained about 70% and 50%, of this activity, respectively. The GUS-TR3337 activity decreased significantly following at heat treatment 90uC for 10 min and was undetectable if subjected to 90uC or higher temperatures for 30 min or longer ( Figure 5).

Discussion
Various reporter genes encoding GFP, GAL, LUC, GUS, or oxalate oxidase (OxO) have been used in transgenic plant research or crop development, and have been assessed for the efficiency, scientific application and commercialization. Many of these genes have specific limitations or have not been sufficiently tested to merit their widespread use [1]. There are some reports on improved reporter genes that allow easy detection or a wide range of applications and that are generated by directing their evolution in the laboratory. Crameri and his colleagues used DNA shuffling to obtain GFP variants with 40 times increased fluorescence intensity [42]. A mutant GFP with stronger fluorescence intensity was selected by directed evolution in conjunction with a functional salvage screen [43].
The GUS enzyme encoded by gusA gene of E. coli is still to date the most widely used reporter gene in plants. Some limitations of using GUS as a reporter are the activity assay that is destructive to the plants and the false positive events due to endogenous GUS or GUS-like enzymes activity in some plants [44]. Hu et al. reported that there were intrinsic GUS-like activities in some seed plants [13]. Some novel GUS or GUS-like genes were also isolated and characterized from plants, such as Scutellaria baicalensis Georgi [45]. GUS-like enzymes were known to be present in microorganisms [2,3,46,47], animals [48,49] and plants [11,13,36,39,50], such as Scutellaria baicalensis [51], rye [52], and maize [53]. Sudan and his colleagues demonstrated that GUS is ubiquitously present in plants [36]. Over the past three decades, numerous reports have illustrated that there was GUS background activity in different plant species, including some model plants, such as Arabidopsis thaliana, Oryza sativa, Nicotiana tabacum and Zea mays [36,38,54,55,56]. Therefore, efforts have been focused on the suppression of the background activity [57,58,59]. As reported by Kosugi et al., the addition of methanol to 20% of the final volume during GUS staining, or during fluorimetric assays, eliminates the endogenous activities found in many plants [59]. The methanol appears to have no effect on the bacterial GUS encoded by the gusA from E. coli. However, methanol is not always effective in suppressing the endogenous GUS activities, and moreover, it is harmful to the experimentalists. The E. coli GUS system is a good model system for the study of directed evolution in vitro. Matsumura et al. have directed the evolution of a GUS variant that is significantly more resistant to both glutaraldehyde and formaldehyde than the wild-type enzyme [40]. This GUS variant could resolve the problem of loss of GUS activity in transgenic plants during tissue fixation by glutaraldehyde or formaldehyde. After four cycles of screening the best variant was more active than the wild-type enzyme, and retained function at 70uC, whereas the wild-type enzyme lost activity at 65uC [41]. DNA shuffling (sexual recombination) and a histochemical screen have also been used to direct the evolution of E. coli GUS variants with improved b-glucuronidase activity [60]. The same model evolutionary system was employed to test the efficiency of several other techniques [61]. One of the GUS variants was obtained by screening for increased p-nitrophenyl-bd-xylopyranoside (pNP-xyl) activity [62]. Although there have been a few studies on directed evolution of GUS to obtain an improved reporter, none of them have been in transgenic plants.
We obtained a temperature resistant gusA variant gus-tr3337 (GenBank No. DQ513151) by employing a high throughput system for directed evolution in vitro [32,33]. The mutant enzyme maintained its activity even when the nitrocellulose filter containing the variant colony was heated at 100uC for 30 min.
Sequencing of the gus-tr3337 revealed six mutations: Q493R, T509A, M532T, N550S, G559S and N566S that are present in GUS-TR3337, indicating that these were the key amino acids needed to confer its high thermostability. The purified GUS-TR3337 protein retained most of its activity when heated at 80uC for 10 min, while the wild-type GUS protein lost its activity completely after 10 min at 70uC. With pNPG as a substrate, the purified GUS-TR3337 protein retained 75% or 48% of its activity when heated at 80uC for 10 min or 30 min, respectively [33]. In the present study, with 4-MUG as substrate, transgenic Arabidopsis plants hosting the gus-tr3337 gene retained about 70% and 50% of GUS activity when heated at 80uC for 10 min or 30 min, respectively. With pNPG as substrate, a single activity peak had been recorded at pH 6.5 at 37uC with the purified GUS-WT and GUS-TR3337 enzymes [33]. Similar results were obtained in the present study on GUS enzymes in extracts from transgenic Arabidopsis plants overexpressing the gus-tr3337 gene and the gus-wt gene. The apparent optimal temperature for GUS-WT and GUS-TR3337 tested at pH 6.5 was 45uC and 65uC, respectively with pNPG as substrate; while temperature optima for the two enzymes expressed in Arabidopsis plants, as tested with 4-MUG as a substrate at pH 7.0 was 45uC and 60uC respectively. Thus, the apparent temperature optimum of GUS-TR3337 expressed in transgenic plants was 5uC lower than that of purified enzyme, under the conditions of the assays. We also repeated the characterization (determination of the optimal pH, temperature and of residual activity after heat treatment) of the purified enzymes of GUS-TR3337 and GUS-WT by comparing pNPG and MUG as substrates and found that the two substrates yield similar results ( Figures S3 and S4).
Although the GUS system is the most popular transgenic reporter gene for plants [1], there were some limitations of the use GUS system as a reporter. Other potential problems with GUS, as discussed by Taylor [44], include its false positive, particularly in those plants with endogenous high endogenous GUS or GUS like activity that interferes with the histochemical detection of transgenic GUS. In the present study, transgenic Arabidopsis plants with the gus-tr3337 gene (GenBank No. DQ513152) retained GUS enzyme activity in plant extracts as well in vivo even after being heated at 80uC for up to 30 min, while plants with the gus-wt gene lost both in vitro and in vivo enzyme activity almost completely after 10 min at 60uC. Based on the staining with 1 mg/mL X-GLUC the apparent temperature optimum for GUS-TR3337 activity was found to be 60uC. The transgenic Arabidopsis plant hosting the gus-tr3337 gene could be easily staining with X-GLUC after heated at 80uC for 10 min to 30 min. The thermostability of GUS-TR3337 can be exploited to minimize the recovery of false positives. Some limitations of GUS as a reporter could be overcome by using this thermostable GUS gene (gus-tr3337). Staining with X-GLUC after  heat treatment at 70uC for 30 min or at 80u for 10-30 min will allow to distinguish the GUS activity associated to the transgenic gus-tr3337 expression from the endogenous one, responsible of false positives, in the plants where the endogenous Gus or GUSlike activity is not thermostable. In conclusion, the use of gus-tr3337 transgene can enhance the precision, sensitivity, and reliability of gus gene as a reporter. Figure S1 Schematic diagram of the vectors used in this study. The original (gus-wt) and mutant (gus-tr3337) genes was cloned into the binary vector pYF7716, at the Bam HI and Sac I restriction enzyme sites, under the control of double CaMV 35S promoter (D35S). For steady transmission of gus-wt and mutant gus-tr3337, two scaffold attachment regions (SAR) were fused upstream of the D35S promoter and downstream of the Nos-Terminator (Nos-T).