Functional characterization of glycine oxidase from Bacillus licheniformis in Escherichia coli and transgenic plants

To find glycine oxidase genes that can be applied to the breeding of glyphosate resistant crops. The glycine oxidase (GO, EC 1.4.3.19) gene (GenBank No: KC831746) from Bacillus licheniformis (B. licheniformis) was chemically synthesized and transformed into glyphosate-sensitive Escherichia coli (E. coli). The GO gene was transformed into Arabidopsis and rice through Agrobacterium-mediated transformation. The test results confirmed that transgenic plants containing GO genes are more resistant to glyphosate than wild-type plants. On solid Murashige and Skoog (MS) (Murashige and Skoog1962 ) medium containing 200 µM glyphosate, transgenic Arabidopsis thaliana grew normally, while wild-type plants were stunted and root growth was restricted. In a solution containing 500 µM glyphosate, wild-type rice showed severe yellowing, while transgenic rice grew normally. In addition, when sprayed with 10 mM glyphosate solution, wild-type rice withered and died, while transgenic rice grew well. The function of GO gene in glyphosate resistance and the application value of GO gene in the cultivation of glyphosate-resistant crops is proved. The glycine oxidase gene from B. licheniformis enhances the resistance of E. coli, Arabidopsis and rice to glyphosate.


Introduction
Glyphosate is the most widely used and the most used herbicide. According to statistics in 2015, planting areas of glyphosate-resistant crops accounted for half of the global genetically modified crops (Du 2020). Most of the grown transgenic glyphosate-resistant crops get resistance to overexpressing EPSPS isolated from Agrobacterium sp. CP4, or a variant of EPSPS, as well as a mutated maize EPSPS. The theoretical disadvantage of this approach is that glyphosate remains of the plant and crop yield (Li and Cao 2018). Neither of the glyphosate detoxification mechanisms have been shown to occur to higher plants to a significant degree (Pollegioni et al. 2011).
Another strategy is to introduce glyphosatedegrading enzymes to degrade glyphosate into non-toxic products. GO was cloned from Bacillus subtilis (Nishiya and Imanaka 1998). GO can convert glyphosate into amino methyl phosphonic acid (AMPA) and glyoxylate by cutting the carbon-nitrogen bond. These two substances are much less toxic than glyphosate. In a previous work, Pedotti et al (2009) improved the catalytic ability of the GO gene to the substrate glyphosate through rational protein designed. The research results showed that the GO gene obtained by artificial evolution has good catalytic activity against glyphosate, which provides a mechanism for glyphosate resistance. Zhang et al. (2013) cloned the glycine oxidase gene from Bacillus cereus, and directed evolution to obtain a mutant with higher catalytic efficiency for glyphosate. Nicolia et al. (2014) transferred a plant-optimized GO variant of Bacillus subtitlis into alfalfa. Transgenic alfalfa plants showed moderate resistance to glyphosate. Yao et al. (2015) improved the activity of B3S1 (GO from Bacillus cereus) through DNA shuffling, and the specific constant of mutant B4S7 increased by 3.9 times. In conclusion, GO-mediated glyphosate resistance is based on herbicide degradation and has the potential to reduce glyphosate phytotoxicity in crops. Zhang et al. (2016) cloned and characterized the GO gene (GenBank No: KC831746) from B. licheniformis. The kinetic parameters on glycine (K m , K cat , and K cat /K m ) of the GO were 0.9 mM, 0.31 s −1 , and 0.03 mM −1 s −1 . The kinetic parameters of glyphosate (K m , K cat, and K cat /K m ) of GO were 11.22 mM, 0.08 s −1 , and 0.01 mM −1 s −1 , respectively. It showed great potential for degrading activity towards glyphosate. We chemically synthesized the wild-type GO gene from B. licheniformis and transformed it into glyphosate-susceptible E. coli to detect GO-imparted resistance. Furthermore, to evaluate its potential, we also transformed the gene into Arabidopsis and rice and demonstrated that transgenic plants exhibit significant glyphosate resistance when compared with the wild type.

Design and chemical synthesis of GO gene of B. licheniformis
According to the sequence in NCBI, the GO gene has been synthesized by the continuous polymerase chain reaction (PCR) method . PCR was carried out as described by Tian et al. (2011) (see primers in Supplementary Table 1). The amplified fragment was digested by BamHI and SacI, cloned into Simple pMD-18, and sequenced . Errors in the synthetic gene were corrected by the overlap extension PCR method (Xiong et al.2006).

Glyphosate sensitivity assays of GO gene in E. coli
The wild GO gene was inserted into the BamHI-to-SacI site of p251. Recombinant p251 was transformed into competent DH5α to determine the gene resistance to glyphosate. Empty p251 was transformed into competent DH5α as a control. The positive monoclonal of two transformants were screened by 100 mg Ampicillin l −1 . Positive monoclonal were cultured in LB medium with 100 mg Ampicillin l −1 to OD 660 of 0.3 and diluted 10, 100, and 1000 times, respectively. Then 2 μl of each transformant was placed on M9 solid medium containing 0 and 200 mM glyphosate, at 37 ℃ for 16 h. At the same time, two kinds of positive transformants with the same amount were transferred into fresh liquid M9 medium with 0 and 200 glyphosate, shaking at 37 °C, and OD 660 values were determined every 2 h (Sun et al. 2005).

Plant expression vector construction and plant transformation
The construction and transformation of Arabidopsis expression vector refer to the description of Xu et al. (2010) and made some modifications. A chloroplast transit sequence was added to the constructs to ensure the GO gene was targeted to the chloroplast. The DNA encoding the chloroplast transit peptide (TSP) was amplified according to Klee et al. (1987). The DNA fragment was amplified (see primers in Supplementary Table 2). The recombinant plasmid, D35S:TSP:GO:Nos was transformed into Agrobacterium GV3101 by electroporation, and then transformed into Arabidopsis by a floral dip method described by Zhang et al. (2006).
Construction of rice expression vector is the same as above. The recombinant plasmid was electroporated into Agrobacterium EHA105. Transgenic plants were obtained according to the method of Tian et al. (2011). The constructs were introduced into rice using callus derived mature embryos. After two rounds of screening with N6 medium containing 30 and 42 mg hygromycin l −1 , the embryos were transferred to MS medium containing 2.5 mg 6-BA l −1 and 0.2 mg NAA l −1 to differentiate into seedlings. The 3-4 cm rice seedlings were transferred to rooting culture medium (42 mg hygromycin l −1 , 0.2 mg NAA l −1 , 1/2 MS). After the rice seedlings completely rooted, they were planted in the greenhouse (25 °C, 16 h light/8 h dark cycle).

Selection and RT-PCR detection of transgenic plants
Arabidopsis T0 generation seeds were sterilized with 2.5% bleach and spread on MS medium plates containing 30 mg hygromycin l −1 . The whole process of T0 generation rice was accompanied by hygromycin screening. The obtained transgenic plants were verified by RT-PCR. The cDNA of the GO gene and RT-PCR was performed as Zhu et al. described previously (2012). Total RNA was extracted from T2-generation transgenic plants. The first cDNA strand was synthesized in 50 μl reaction solution using 10 ng RNA as a template. To improve the reliability of RT-PCR, the Arabidopsis actin gene, AtAc2 (GenBank No: NM12764) and rice actin gene (Gen-Bank No: X16280) (see primers in Supplementary  Table 3) were used as the internal standard. The GO gene fragment (300 bp) was amplified by specific primers (see primers in Supplementary Table 3). The positive Arabidopsis and rice plants were carried out for two more generations in order to obtain homozygous transgenic plants.
The glyphosate resistance of transgenic plants Glyphosate resistance of Arabidopsis analysis was performed as Han et al.(2015). The wild-type (WT) and T2 transgenic seeds were sterilized with 2.5% bleach and then sowed on solid MS medium with 0 and 200 µM glyphosate, respectively, and grown for 12d in a controlled-environment chamber (22 °C, 16 h light/8 h dark cycle). Seeds of wild-type Arabidopsis and transgenic Arabidopsis were cultured in pots containing a mixture of vermiculite, peat moss, and perlite (18:6:1, v/v) in a controlled-environment chamber (22 °C, 16 h light/8 h dark cycle) for 4 weeks. The four-week-old seedlings were sprayed with 10 mM glyphosate, doses of 30 ml m −2 every 3 days. One day after second spraying, when the leaves of WT Arabidopsis showed a slight yellowing, the total chlorophyll content was measured as Lichtenthaler et al. described previously (1987).
After accelerating germination at 37 ℃ for 24 h in the dark, the wild type and transgenic rice seedlings with uniform growth vigor were transferred into the solution containing 0, 250, and 500 µM glyphosate for 7d (25 °C, 16 h light/8 h dark cycle). For glyphosate-sprayed transgenic rice, wild-type and transgenic plant seeds were cultured in pots containing a mixture of vermiculite, peat moss, and perlite (18:6:1, v/v) in a controlled-environment chamber (25 °C, 16 h light/8 h dark cycle). After 3 weeks, the seedlings up to 12-15 cm in height were sprayed with 10 mM glyphosate at a dose of 70 mLm −2 .

Chemical synthesis and glyphosate sensitivity test of E. coli
The gene was chemically synthesized using primers according to the original sequence. PCR amplification of the entire open reading frame of GO produced a single band of 1110 bp, encoding 369 amino acids.
DH5α cells transformed with empty p251 plasmid were completely inhibited on M9 plates containing 200 mM glyphosate. The DH5α cells transformed with p251-GO plasmid grew well on 200 mM glyphosate plates and were much less inhibited (Fig. 1).
The growth curves of E. coli DH5α containing p251 or p251-GO are shown in Fig. 2. After 24 h of culture, the growth of cells containing p251 was strongly inhibited by 200 mM glyphosate (> 95%), while the OD value of the culture containing p251-GO was about 37% of the control values without glyphosate.

Plant transformation and transgenic plant selection
The expression cassette composed of CaMV35S double promoter, chloroplast transport sequence, and GO coding region was cloned into the T-DNA region of the binary vector (Fig. 3a). Transgenic Arabidopsis and rice plants were obtained from MS medium containing 30 and 42 mg/l hygromycin, respectively, and verified by RT-PCR analysis. RT-PCR was used to analyze the gene expression of three transgenic Arabidopsis lines (Go1, Go3, Go7) and three transgenic rice lines (Gor3, Gor7, Gor8). The transcription level of GO gene was detected by RT-PCR. In Growth of E. coli harboring p251-GO and empty p251 in liquid M9 minimal medium supplemented with glyphosate at 0, 200 mM concentrations. p251, DH5α with empty p251 plasmids; p251-GO, DH5α with p251-GO plasmids. The error bars mean ± SD of three repetitions the T2 generation, the expression of GO mRNA in the three overexpression lines was high. In addition, the transcript was not detected in wild-type plants (Fig. 3b, c). Under normal growth conditions, GO transfer had no significant effect on plant growth and development.

Glyphosate resistance analysis of transgenic Arabidopsis plants
The wild-type and transgenic plants grew consistently on the control medium without glyphosate. On a medium containing 200 µM glyphosate, the transgenic Arabidopsis developed normally, while wildtype Arabidopsis developed slowly and root growth was limited ( Fig. 4; Fig. 5). There were also differences in root length between transgenic lines. This may be due to the different expression levels of GO gene in plants (Fig. 4). Previous studies have shown that sublethal concentrations of glyphosate caused yellowing symptoms in transgenic tobacco leaves expressing OsEPSPs gene (Zhou, 2006). Therefore, the total chlorophyll content of all lines was measured one day after the second spraying of 10 mM glyphosate (WT Arabidopsis leaves showed slight yellowing). It was found that the chlorophyll content of transgenic lines was higher than that of wildtype Arabidopsis (Fig. 6). These results indicated Scale bars 1 cm that the transgenic Arabidopsis showed higher resistance to glyphosate than the wild-type.

Glyphosate resistance analysis of transgenic rice plants
Three overexpression lines (named Gor3, Gor7, and Gor8) were tested for glyphosate-resistant experiments. It was found that WT and transgenic plants grew consistently in glyphosate-free solution. In the Origin. *means P < 0.05, **means P < 0.01. Error bars represented the SD of the mean calculated for three replicates Fig. 6 Analysis of chlorophyll content after spraying glyphosate. Concentration of spraying glyphosate, 10 mM; WT, wild-type; Go1, Go3, Go7, transgenic Arabidopsis lines. The asterisk on the top of the bar indicates different significances according to one-way ANOVA using Origin. *means P < 0.05, **means P < 0.01. Error bars represented the SD of the mean calculated for three replicates At 500 µM glyphosate concentration, wildtype rice almost died due to severe yellowing. The plant height of transgenic lines was inhibited but could grow normally (Fig. 7).
As shown in Fig. 8, 7 days after spraying with 10 mM glyphosate on the leaves of three-week-old rice plants, most of the leaves of the wild-type rice plants wilted, became excessively dehydrated and died. Only the leaves of some transgenic plants and others grew well and had normal morphology.

Discussion
The safety of herbicide-resistant transgenic crops has also attracted attention while bringing huge benefits to agricultural production (Vicini et al. 2019). The resistance genes of existing commercial crops are mainly existing commercial glyphosate-resistant crops, using microorganisms (Agrobacterium sp. CP4) or not subject to glyphosate (Ye et al. 2010) in the form of EPSPs mutations (TIPS) concerns. This mechanism does not eliminate or degrade glyphosate. Therefore, the search for new sources of genes capable of degrading glyphosate will become very meaningful. The GO gene has entered people's vision because its substrate has a similar molecular structure to glyphosate. Unfortunately, there is no endogenous GO in plants, so it is necessary to transfer exogenous GO genes to achieve the purpose. The GO gene can degrade glyphosate into AMPA and glyoxylic acid to reduce glyphosate residue. For example, Lu (2011) obtained a glycine oxidase gene (GOA, Genbank accession number: GU479460) by screening herbicide-degrading genes from a metagenomic library. The gene has good substrate affinity and substrate selectivity for glyphosate. Zhang et al (2016) cloned glycine oxidase from B. licheniformis and improved the catalytic activity of glyphosate by error-prone PCR and two rounds of DNA shuffling. In addition, glyphosate oxidase was obtained by directed evolution of glycine oxidase and coupled with spore-luminol, improving the sensitivity and high selectivity of glyphosate detection (Qin et al 2020). These studies have shown the degradation efficiency of glyphosate by GO gene. To make GO gene more glyphosate tolerant in plants, we chemically synthesized GO gene from B. licheniformis. On this basis, the GO gene was chemically synthesized according to plant codon preference and transferred into Arabidopsis and rice by Agrobacterium. The results confirmed that the obtained transgenic lines showed high resistance to glyphosate, and the chlorophyll content of transgenic Arabidopsis was significantly higher than that of wild-type Arabidopsis (Fig. 6). Particularly at late developmental stages, transgenic Arabidopsis had longer roots, stronger leaves, and better development (Figs. 4,5). In addition, transgenic rice has the same performance. Under high-concentration glyphosate (10 mM), most of the wild-type rice withered and died, while the leaves of transgenic rice planted grew well and developed normally (Figs. 7, 8). These results indicate the function of the GO gene for