Optimizing plant transporter expression in Xenopus oocytes

Background Rapid improvements in DNA synthesis technology are revolutionizing gene cloning and the characterization of their encoded proteins. Xenopus laevis oocytes are a commonly used heterologous system for the expression and functional characterization of membrane proteins. For many plant proteins, particularly transporters, low levels of expression can limit functional activity in these cells making it difficult to characterize the protein. Improvements in synthetic DNA technology now make it quick, easy and relatively cheap to optimize the codon usage of plant cDNAs for Xenopus. We have tested if this optimization process can improve the functional activity of a two-component plant nitrate transporter assayed in oocytes. Results We used the generally available software (http://www.kazusa.or.jp/codon/; http://genomes.urv.es/OPTIMIZER/) to predict a DNA sequence for the plant gene that is better suited for Xenopus laevis. Rice OsNAR2.1 and OsNRT2.3a DNA optimized sequences were commercially synthesized for Xenopus expression. The template DNA was used to synthesize cRNA using a commercially available kit. Oocytes were injected with cRNA mixture of optimized and original OsNAR2.1 and OsNRT2.3a. Oocytes injected with cRNA obtained from using the optimized DNA template could accumulate significantly more NO3- than the original genes after 16 h incubation in 0.5 mM Na15NO3. Two-electrode voltage clamp analysis of the oocytes confirmed that the codon optimized template resulted in significantly larger currents when compared with the original rice cDNA. Conclusion The functional activity of a rice high affinity nitrate transporter in oocytes was improved by DNA codon optimization of the genes. This methodology offers the prospect for improved expression and better subsequent functional characterization of plant proteins in the Xenopus oocyte system.

Although all organisms generally share the same genetic code, each genus has evolved a slightly different pattern of codon usage. Heterologous protein expression in a foreign host may be diminished by factors such as biased codon usage, GC content and repeat sequences. To overcome these limitations, codon optimization can be used to enhance gene expression in various host cells. Heterologous synthetic genes with codon optimization showed increased expression levels in various organisms including E. coli [24,25], yeast [26] and mammalian cells [27,28]. For many plant transporters expressed in Xenopus oocytes, the low levels of expression can often limit the functional assay, making the detailed characterization of the protein difficult. In the past, it was speculated that differing codon bias may explain the very low levels of expression of some plant proteins in oocytes [29]. Improvements in DNA synthesis technology have enabled the technique to be used for costeffective gene cloning. Commercial suppliers make it possible to obtain the synthetic DNA with codon optimization in just a few weeks. In this study, DNA of the rice genes OsNAR2.1 and OsNRT2.3a were codon optimized and synthesized for oocyte expression. The cRNA of OsNAR2.1 and OsNRT2.3a were then synthesized using a commercially available kit. We compared how this process may improve the functional activity of plant nitrate transporter proteins expressed in oocytes. The nitrate transport activity was assayed using 15 N-enriched nitrate uptake and the two-electrode voltage clamp technique.

Results and discussion
Codon optimization of OsNAR2.1 and OsNRT2.3a There are some general rules that emerge from the analysis of the preferred codons in Xenopus and these can be used to optimize a gene sequence for expression in oocytes [30]. Using the codon usage bias software for Xenopus (http://www.kazusa.or.jp/codon/; http://genomes. urv.es/OPTIMIZER/), the DNA gene sequence was optimized for OsNAR2.1 (LOC_Os02g38230) and OsNRT2.3a (LOC_Os01g50820) and the resulting DNA sequences were synthesized by the Genescript Company and named syn-OsNAR2.1 and syn-OsNRT2.3a. After software optimization, the predicted GC content of syn-OsNAR2.1 and syn-OsNRT2.3a was adjusted from 72.0 to 52.6% and 67.2 to 49.0% respectively, when compared with the original genes (Table 1). This change now makes the plant genes synthetic DNA much closer to the typical 50% GC content found in Xenopus [31]. For both synthetic DNAs the melting temperature (T m ) was decreased and the number of repeat sequences was decreased in syn-OsNAR2.1 (see Table 1). Sequence alignment of the open reading frames showed that syn-OsNAR2.1 and syn-OsNRT2.3a shared 73% and 74% identity with the original genes ( Figure 1), but the amino acid sequences did not change after optimization (see Additional file 1).

Nitrate uptake of oocytes
The original and synthetic (optimized) OsNAR2.1 and OsNRT2.3a were subcloned in to expression vector pT7Ts [30] and then used as template to synthesize mRNA. Mixed mRNA of either synthetic genes (syn-OsNAR2.1: syn-OsNRT2.3a) or the original genes (ori-OsNAR2.1: ori-OsNRT2.3a) were injected into oocytes. Both RNA mixes were injected at the same ratio (OsNAR2.1:OsNRT2.3a, 25:50 ng). We used a colorimetric method to assay the amount of nitrate accumulated inside the oocyte. After 16 h incubation in MBS containing 0.5 mM NaNO 3 , oocytes injected with mRNA of synthetic genes showed increased NO 3 uptake when compared with the original genes ( Figure 2). Similar results were obtained in 5 mM NaNO 3 (Additional file 2) . These data did not show a significant difference between the original genes and water injected oocytes. In another set of experiments, injected oocytes were incubated in MBS containing 0.5 mM Na 15 NO 3 for 8 and 16 h. Compared to original genes, 15 NO 3 uptake of single oocyte injected with synthetic genes were greatly enhanced after 8 h and 16 h incubation ( Figure 3). Individual oocytes injected with synthetic gene mRNAs generally showed much greater 15 NO 3 uptake  Figure 4A, B). The longer incubation time resulted in more nitrate accumulation also perhaps a greater concentration of transporter protein in the membrane has developed after 16 h. Statistical analysis showed that 100% of oocytes injected with synthetic DNA had significantly increased 15 NO 3 uptake after 8 h and 16 h, while the equivalent figure was only 14% and 17% for the original plant template relative to water-injected controls ( Figure 4). Presumably, the low percentage of nitrate transporter activity in oocytes injected with RNA from the plant DNA template can explain the results shown in Figure 2, where there was no significant difference between the total nitrate uptake of original genes-injected oocytes and waterinjected ones in an 8 h uptake experiment.

Electrophysiological analyses of oocytes
Two-electrode voltage clamp analysis was performed to record the voltage-current relationships of oocytes injected with mRNA [2,3]. After 48 h mRNA injected oocytes were treated with 0.5 mM NaNO 3 . Under these conditions when the plasma membrane voltage was clamped, in this example the nitrate-elicited currents of an oocyte injected with the synthetic genes was twice as large as an oocyte injected with RNA produced from the original plant DNA template ( Figure 5). The electrophysiological measurements confirmed the accumulated nitrate ( Figure 2) and 15 N-nitrate influx (Figure 3) data in showing larger nitrate-elicited currents in oocytes injected with RNA made using the synthetic optimized DNA template. The 15 N influx data was the average of 20-25 oocytes and showed an 8-fold advantage of using  Table 1 more specific sequence differences are listed. the optimized DNA. These data demonstrate the significant methodological advantage of using a template DNA that has been optimized for Xenopus expression.

Enhanced transporter activity in oocytes
The expression of foreign proteins in oocytes has long been known to be improved by the inclusion of a polyA tail and the use of expression vectors that include Xenopus globin flanking UTR (untranslated region) sequence alongside the foreign DNA [1,29,30]. The polyA tail is recognized to improve mRNA stability and lifetime in the oocyte thereby improving the production of a foreign protein. The frog globin flanking UTR sequence is thought to make the heterologous cDNA more Xenopus-like and therefore thought to improve translation [29,30]. Similarly the mRNA produced from the synthetic DNA is more like Xenopus message and this has resulted in improved translational efficiency in the oocyte. For OsNAR2.1 the removal of some repeat sequence (Table 1) may also have given better translation of the foreign protein. Making the GC content more like the 50% found in Xenopus [31] is likely to improve expression of plant genes. In Arabidopsis the GC content was reported as 44%, on the other hand in maize, rice and barley the figure was higher at >60% [32]. Mammals usually have around 44% GC in their coding sequence and experimental work directly comparing low-GC genes with their high-GC counterparts showed 100-fold greater expression in the GC-rich genes [33]. This study also   showed that the mRNA degradation rate was independent of the GC content.
Together these data clearly show the methodological advantage for plant genes of using a synthetic template that has codon usage more like that found in Xenopus. It is widely accepted that optimized codons help to achieve faster translation rates and high accuracy in bacteria, yeast and mammalian cells [24][25][26][27][28]34] and we now show this has advantages for plant genes in Xenopus oocytes too.

Conclusion
In this study, rice nitrate transporter OsNAR2.1 and OsNRT2.3a were codon optimized for Xenopus laevis. Nitrate transport activity was analyzed and compared between oocytes injected with different sources of template DNA. The optimization changes the DNA, but not the protein sequence. Compared with the original plant genes, oocytes injected with optimized genes had increased nitrate uptake and larger currents in electrophysiological analyses suggesting that there was an increased level of protein expression. Taken together, these data show that the codon optimized template can give much improved expression and therefore provides a big advantage when aiming to functionally characterize a plant transporter protein in the Xenopus oocyte system. Although this may not be the case for all plant transporter genes the relatively cheap cost of DNA synthesis now makes this worthwhile when using oocyte expression.

Methods
Cloning and mRNA synthesis of OsNAR2.1 and OsNRT2.3a OsNAR2.1 and OsNRT2.3a were codon optimized and synthesized by the Genescript Company. cDNAs were then subcloned into the BglII and SpeI sites of the oocyte expression vector pT7TS [35] using a directional cloning method. Original OsNAR2.1 and OsNRT2.3a construct are as described previously [22]. Plasmid was linearized by BamH I (Roche) and purified by PCR purification kit (QIAGEN). The mRNA synthesis kit (mMESSAGE mMA-CHINE® T7 Kit, Ambion) was used to synthesized the mRNA of all genes. A compressed air system (Harvard) was used for injection [1]. Glass tips were calibrated for injection using known volumes and 1 μl mRNA mixtures (0.5 μg OsNAR2.1 and 1 μg OsNRT2.3a in total) were used to inject around 22-25 oocytes. Thus mRNA mixtures were injected as 25 ng: 50 ng, OsNAR2.1: OsNRT2.3a, and this weight ratio was chosen to reflect the differing molecular sizes and give similar molecular ratios [22,23]. Oocytes from three different frogs were used for the data shown.

NO 3 accumulation and 15 N-NO 3 uptake in oocytes
After injection, the oocytes were incubated for 2 days in NO 3 free MBS solution (88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO 3 , 0.71 mM CaCl 2 , 0.82 mM MgSO 4 and 15 mM HEPES, pH 7.4). The solution contains 10 g/ml sodium penicillin and 10 g/ml streptomycin sulphate. For NO 3 measurement, oocytes were incubated in MBS solution containing 0.5 mM NaNO 3 at 18°C for 16 h. After incubation, oocytes were washed with NO 3 free MBS for four times. Four oocytes were collected as one sample in 1.5 mL tube, and 100 μl H 2 O was added into the tube. Lyses the oocytes and centrifuge the tube at 13000 rpm. The supernatants (40 μl) were collected for NO 3 assay using the kit (Nitrate/Nitrite Colorimetric Assay kit, Cayman).
For 15 N-NO 3 measurement, oocytes were incubated in MBS solution containing 0.5 mM Na 15 NO 3 with a 99% atom excess of 15 N for 8 h and 16 h. oocytes were washed four times with ice-cold 0.5 mM NaNO 3 MBS. Single oocyte was transferred to an empty tin capsule and then dried at 60°C for one week. Analysis for total 15 N content using a continuous-flow isotope ratio mass spectrometer coupled with a carbon nitrogen elemental analyzer (ANCA-GSLMS; PDZ Europa). The delta-15 N was calculated as described previously [4].

Two-electrode voltage clamp analysis
The nitrate-elicited currents were recorded in oocyte using two-electrode voltage clamp method (pClamp 10.2, Axon). The oocytes were incubated in nitrate-free MBS and then treated with MBS containing 0.5 mM sodium nitrate. Membrane potential of oocytes was pulsed from 0 to −160 mV with 20 mV incremental steps. The currents were recorded to obtain current-voltage curves [2][3][4]. These curves were recorded from oocytes injected with water (blue), synthetic (red) and original (green) genes. All the oocytes treated with 0.5 mM nitrate at pH 7.4. Results were average and SE values and obtained in three different oocytes from the same frog.