The rice annexin gene OsAnn5 is a positive regulator of cold stress tolerance at the seedling stage

Annexins exist widely in plants as multigene families and play critical roles in stress responses and a range of cellular processes. In this study, we report on the cloning and functional characterization of the rice annexin gene OsAnn5. We found that the expression of OsAnn5 was induced by cold stress treatment at the seedling stage of rice. GUS staining assay indicated that the expression of OsAnn5 was non tissue-specic and was detected in almost all rice tissues. Subcellular localization indicated that OsAnn5-GFP (green uorescent protein) signals were found in the endoplasmic reticulum apparatus. Compared with wild type rice, overexpression of OsAnn5 signicantly increased survival rates at the seedling stage under cold stress, while knocking out OsAnn5 using the CRISPR/Cas9 (clustered regularly interspaced short palindromic repeats/CRISPR associated proteins) mediated genome editing resulted in sensitivity to cold treatments. These results indicate that OsAnn5 is a positive regulator of cold stress tolerance at the seedling stage. OsAnn5 knock out (KO) line and also constructed OE, OsAnn5pro::GUS and OsAnn5-GFP lines. We found that the OE lines enhanced cold tolerance in rice, whereas the KO lines were more sensitive to cold stress. activity the nal phenotypic determination: once off-target activity takes place, it becomes dicult to determine whether phenotypic change is due to target gene knockout or off-target activity. it is to evaluate potential off-target loci when we carry out gene knockout using a CRISPR/Cas9-mediated approach. Off-target mutations caused by the CRISPR system can be minimized by choosing target sequences that have reduced numbers of off-targets. is a convenient and integrated toolkit by which we can expedite all experimental designs and analyses of mutations for CRISPR/Cas9 genome editing in plants, and it provides a set of powerful tools for prediction of off-target sites. In this study, we evaluated ve candidate off-targets, including two in the exons of OsAnn5 using the CRISPR-GE software package. We found no evidence for off-targeting phenomena in the candidate sites. These results support the reliability of our identication of the cold tolerance phenotypes in mutants. our results demonstrated that based on the electrical conductivity and survival ratio tests, T 1 mutant lines from two T 0 biallelic mutants showed decreased cold tolerance compared with the Taipei309 WT variety. Additionally, overexpression of OsAnn5 improved tolerance to cold stress in rice. These results indicate that the rice annexin gene OsAnn5 is a positive regulator of cold stress tolerance at the seedling stage. OsAnn5 thus becomes only the second rice annexin gene reported to be involved in cold tolerance at the seedling stage, following previous reports of a similar role for the annexin gene OsAnn3. These results expand our understanding of the complex mechanisms of annexin response to cold stress in rice. Genetic engineering using annexin genes might offer a new and excellent platform to develop rice cold resistance breeding.


Background
Abiotic stresses in the environment can disadvantageously affect the normal growth, development and yield of crops. Because of frequent climate abnormalities and inappropriate agricultural management strategies, abiotic stresses have become a major challenge threatening global agricultural production and development. Plant damage from abiotic stresses is mainly caused by the loss of cell homeostasis leading to cell death (Huang et al., 2012;Rengel et al., 2012). In order to maintain the stability of the cell structure and function and survive under adverse conditions, plants have evolved a number of adaptative physiological, biochemical, cellular and molecular responses to abiotic stresses (Bohnert et al., 1995;Browse and Xin, 2001;Chinnusamy et al., 2007). Plants respond to abiotic stresses by regulating the expression of a number of stress-induced genes that may be associated with stress tolerance, transcription regulation or signal transduction (Thomashow 1999;Shinozaki et al., 2003;Nakashima et al., 2009). Transcriptome analysis of four rice genotypes demonstrated that an average of 5975 genes in every genotype, accounting for about 18% of the annotated genes, were differentially expressed under cold stress (Shen et al., 2014). To date, a number of genes have been identi ed that are associated with mechanisms of abiotic stress defense, and annexin genes are an important category of relevant genes .
Annexins are an evolutionarily conserved multigene family of Ca 2+ -dependent phospholipid-binding proteins that occur widely in plants and animals (Rescher and Gerke, 2004;Mortimer et al., 2008;Jami et al., 2012). Sequence analysis has demonstrated that plant annexins harbor motifs or residues related to peroxidase and ATPase/GTPase activity, as well as calcium channel activity (Mortimer et al., 2008), which has also been well demonstrated in subsequent research (Gorecka et Wang et al., 2018). Plant annexins play a role in diverse aspects of plant growth and development, and they are expressed in many tissues from different development stages . Moreover, previous evidence suggests that annexin genes from a range of plant species are transcriptionally activated in response to abiotic stresses. An initial report suggested that the alfalfa annexin gene (AnnMs2) is activated by drought stress, osmotic stress, and ABA treatment (Kovacs et al., 1998). Subsequent evidence suggests that annexins play an important role in other plant abiotic stress responses. For example, AnnAt1 was found to be associated with drought tolerance in Arabidopsis, with more sensitivity to drought stresses in loss-of-function AnnAt1 mutants and improved drought tolerance in gain-of-function mutants (Konopka-Postupolska et al., 2009). AnnAt1 was also found to interact with AnnAt4, such that AnnAt1 and AnnAt4 regulated salt and drought stress tolerance by interacting with each other in a light-dependent manner (Lee et al., 2004;Huh et al. 2010). Overexpression of the annexin gene AtANN8 enhanced salt and dehydration stress tolerance in Arabidopsis (Yadav et al. 2016). In tomato (Solanum pennellii), the annexin gene SpANN2 was found to be involved in drought and salt stress tolerance, with improved growth in SpANN2-overexpression (OE) lines (Ijaz et al., 2017). The cotton annexin gene GhANN1 was also found to be involved in drought and salt stress tolerance (Zhou et al., 2011;Zhang et al., 2015).
Genome sequencing revealed that there are ten annexin genes in rice (Singh et al., 2014), and the functional roles of several of these genes in responding to abiotic stresses have been characterized. The rice annexin gene OsANN1 (Os02g51750) was found to be associated with heat and drought stress response, with more sensitivity to heat and drought stress in RNA interference plants and improved growth in OsANN1-OE lines (Qiao et al., 2014). Similarly, OsANN3 (Os07g0659600) was also con rmed to be a positive regulator of drought stress tolerance in rice in an ABA-dependent manner (Li et al., 2019). We have recently demonstrated that the rice annexin gene OsAnn3 (Os05g0382600) is relevant for cold stress tolerance, with more sensitivity to cold stress when OsAnn3 was knocked out by CRISPR/Cas9-mediated gene modi cation (Shen et al., 2017). This study was the rst report of an annexin gene involved in cold tolerance in rice, despite the fact that low temperature is a common type of stress in the life cycle of rice. In general, the functional and physiological roles of rice annexin genes in responding to cold stress remain unknown.
In the present study we isolate and characterized a putative annexin protein family gene in rice, designated as OsAnn5 (Os06g0221200) [consistent with the nomenclature of Singh et al (Singh et al., 2014)]. We demonstrate that the expression of OsAnn5 increased following low temperature treatment (4 ~ 6 °C for 4 days). We directly tested the role of OsAnn5 by constructing a series of transgenic rice plants; we used CRISPR/Cas9-mediated genome editing to create an OsAnn5 knock out (KO) line and also constructed OE, OsAnn5pro::GUS and OsAnn5-GFP lines. We found that the OE lines enhanced cold tolerance in rice, whereas the KO lines were more sensitive to cold stress.

Plant Materials and Stress Treatment
The rice (Oryza sativa subsp. japonica) cultivar Taipei309 was used in this experiment and was considered as the wild type (WT) control in all experiments. The seeds of Taipei309 and T 1 KO lines from the T 0 biallelic mutant were sterilized and germinated at 37 °C in darkness for 2 days. Then, the seeds were sown in a plastic pot (22 × 17.5 × 7.5 cm) lled with soil in a light incubator, with a 12/12 light/dark cycle at temperatures of 28/25 °C (day/night). The seeds were watered daily using sterile water until the cold stress treatment. When rice seedlings were four weeks old they were transferred to 4 ~ 6 °C for 3 days of cold treatment. With the same light and temperature conditions but not soil and sterile water, the seeds of WT and the T 3 generation OE lines were sown in 9 cm diameter glass petri dishes, and rice seedlings were watered daily using Yoshida solution until four weeks old (until cold treatment). The survival rate and relative electric conductivity were measured as described previously (Shen et al., 2017).

RNA extraction and quantitative Real-Time PCR (qRT-PCR) analysis
Total RNA was extracted from Taipei309 seedling leaves grown under normal (control) conditions or under cold treatment (4 ~ 6 °C for 1 ~ 4 days) using a TransZol Up Reagent Kit according to the manufacturer's protocol (TransGen Biotech, China). The rst stand cDNA synthesis was performed using TransScript® One-Step gDNA Removal and cDNA Synthesis SuperMix Kit (TransGen Biotech, China). SuperReal PreMix Plus (SYBR Green) Kit (TIANGEN, China) was used for qRT-PCR analysis and carried out on a StepOne Real-Time PCR System (Applied Biosystems, USA). Real-time PCR was nished with OsAnn5-F and OsAnn5-R gene-speci c primers (Table S1) as described previously (Shen et al. 2014). The relative expression level was evaluated using means from three biological samples with three technical replicates, and the ampli cation of the ubiquitin gene (Os03g0234200) was used as an internal control for normalizing all data.

β-glucuronidase (GUS) staining
In order to characterize the expression patterns of OsAnn5, we generated OsAnn5 promoter::GUS transgenic rice plants. GUS reporter staining was measured suing histochemical GUS staining (Jefferson et al., 1987). Three positive transgenic rice lines were incubated in 5-bromo-4-chloro-3-indolyl-β-glucuronic acid buffer at 37ºC without any light. After staining, the plant tissues were soaked in 75% ethanol until the chlorophyll ingredient was completely decolorized. Finally, the sample tissues were rinsed with distilled water to remove surface dyes and chlorophyll before being photographed.

Subcellular localization of OsAnn5 protein
The OsAnn5 full-length coding region without stop codon was ampli ed using the primers Ann5-GFP-F/R (Table S1). The PCR product of OsAnn5 was then fused to the GFP N-terminus, and its expression was driven by the CaMV 35S promoter located in the transient expression vector pBWA(V)HS-ccdb-GLosgfp to generate a new construct, pBWA(V)HS-Ann5-GLosgfp. This construct was then co-transformed in rice protoplasts with the marker plasmid harbor red uorescence protein (RFP), and transfected protoplasts were incubated as described previously (Chen et al., 2010). The GFP uorescence was observed using a Nikon C2-ER confocal laser scanning microscope (Nikon, Japan) after 48 h of in ltration.

Construction of OsAnn5 expression vectors
To overexpress OsAnn5, the full-length cDNA was ampli ed from Taipei309 and inserted into the pCAMBIA1301-35S::OsAnn5 vector. To produce a CRISPR/Cas9 expression vector for use in plant gene editing, two targeted sites were designed. DNA oligonucleotides OsAnn5-Oligo1 (24-bp) and OsAnn5-Oligo2 (24-bp) were synthesized according to the targeted site in the third exon of OsAnn5, and DNA oligonucleotides OsAnn5-Oligo3 (24-bp) and OsAnn5-Oligo4 (24-bp) were synthesized on the basis of the targeted site sequence in the fth exon of OsAnn5. After annealing and phosphorylation, they were inserted into BbsI sites of the cloning vector psgR-Cas9-Os (Fig. S1). Then, the targeting single-stranded guide RNA (sgRNA) cassettes and Cas9 in the cloning vector were digested with HindIII and EcoRI, and the fragments were ligated into the same sites of the plant expression vector pSK51 as previously described (Shen et al, 2017). To generate OsAnn5-GFP construct, OsAnn5 full length cDNA was digested by BsaI and Eco31I and then ligated into the pBWA(V)HS-ccdb-GLosgfp vector digested with the same enzymes. To obtain OsAnn5 promoter::GUS construct, about 2 Kb upstream of the OsAnn5 ATG start codon was ampli ed with the primer Ann5pro-F/R and inserted into the KpnI and BglII cloning sites of the vector pCAMBIA1304. The primers used for constructing plasmids are listed in Table S1. The plant expressing vectors were transformed into Taipei309 using agrobacterium tumefaciens-mediated transformation.

Detection of Targeted Gene Mutations
Rice leaf genomic DNA was extracted from the WT rice cultivar Taipei309 and all T 0 transgenic lines modi ed with the CRISPR/Cas9 expression vector using the CTAB method. The sequence segments surrounding the two target sites were ampli ed using high delity DNA polymerase with primer pairs TB-B1-Ann5F/R or TB-B2-Ann5F/R (Table S1). The target site mutations were evaluated by aligning sequencing chromatograms of the T 0 transgenic plants' PCR products with those of the WT rice cultivars. All mutants identi ed by PCR were then subjected to zygosity analysis by means of cloning corresponding PCR products into the pEASY-Blunt Zero Cloning Kit vector (TransGen Biotech, Beijing, China), and 6-8 positive clones from every mutant DNA sample were sent for DNA sequencing.

Off-target Sequence Identi cation
Possible off-target sites were evaluated by comparing the 20-nt gRNA target sequences in OsAnn5 with the whole genomic sequences using a web-based software package, CRISPR-GE (Genome Editing) (http://skl.scau.edu.cn/) (Xie et al., 2017). The e-value threshold was set to 8 automatically because the query sequence (sgRNA) is only 20 nt. When the off-score value is equal or greater than 0.09, sites with the protospacer-adjacent motif (PAM) NGG motif were all considered for analysis regardless of whether they were in exons, introns, or intergenic regions. Speci c primers of possible off-target loci in this experiment are listed in Table S2.

Expression patterns of OsAnn5
The promoter sequence of OsAnn5 was characterized with PlantCARE software http://bioinformatics.psb.ugent.be/webtools/plantcare/html/). We analyzed a 2082 bp DNA sequence upstream of the start codon for OsAnn5 and found that there were several cis-acting elements, including two DRE cores, one MYB recognition site, one MYB-binding site, and one CCAAT-box (MYBHv1 binding site) that have been previously reported to be associated with stress responses (Table S1). To directly evaluate the effect of OsAnn5 in responding to cold stress, qRT-PCR was performed using four week old WT rice seedling leaves under normal conditions (28 °C) or after 4 ~ 6 °C cold treatment for 4 days. Results showed that the transcription levels of OsAnn5 in WT rice followed a low-highhigh-low-low change rule before and after cold stress (Fig. 1). OsAnn5 expression reached the highest level (1.84-fold up-regulated) following 2 days of cold treatment compared to the normal condition (Fig. 1). These ndings suggest that OsAnn5 expression is regulated by cold stress and may be involved in cold tolerance. Additionally, we evaluated the expression patterns of OsAnn5 in various rice tissues using the GUS reporter gene. Results demonstrated the presence of 28 independent positive transgenic rice lines expressing OsAnn5pro::GUS, from which three independent lines were selected to carry out GUS staining experiments. The results of staining indicated that OsAnn5 is expressed in multiple tissues, with the strongest signals found in the node, weaker signals found in the lemma, and staining also found in the embryo, roots, stems, and oral parts (Fig. 2).
Overexpression of OsAnn5 results in enhanced rice cold tolerance To evaluate the function of OsAnn5 in responding to cold stress, T 3 generation OE and WT lines were grown in the same batch of Yoshida solution in 9 cm diameter glass petri dishes for about four weeks, then treated with 4 ~ 6 °C for 3 days, and then returned to the normal growth conditions to recover. After approximately 10 days in the greenhouse following cold treatment, OE lines plants showed better growth, while WT plants had severe wilting and rolling leaves (Fig. 3A). Additionally, OE lines had a higher average survival rate of 39.97% (OE-18, 29.30%; OE-23, 46%; OE-24 44.60%) compared to 6.67% in the WT rice (Fig. 3B). Under the same cold treatment conditions, experiments were carried out to measure the relative electric conductivity of leaves. Results revealed that the relative electric conductivity levels in the three T 3 OE lines were signi cantly lower after cold treatment compared to the WT condition, while they were similar in the three T 3 OE lines and WT rice in non-stress condition (Fig. 3C). These results indicate that overexpression of OsAnn5 in rice can enhance tolerance to cold stress.

OsAnn5-GFP probably localizes to the endoplasmic reticulum apparatus
To determine the speci c subcellular localization of the OsAnn5 protein, rice protoplasts were transformed with the OsAnn5-GFP construct via PEG-mediated transient expression. When OsAnn5-GFP and pBWA(V)HS-ccdb-GLosgfp empty vectors were introduced into the rice protoplasts separately, the distribution of OsAnn5-GFP was more pronounced in endoplasmic reticulum locations compared to the cytosol-localized GFP with empty vector (Fig. 4). We further con rmed the subcellular location of OsAnn5-GFP by expressing it together with the endoplasmic reticulum mcherry marker, with results indicating that the fusion protein was mainly localized to the endoplasmic reticulum apparatus (Fig. 4). Therefore, we concluded that in rice, OsAnn5 is most likely localized to the endoplasmic reticulum apparatus.

Knocking out OsAnn5 resulted in transgenic plants sensitivity to cold stress
One month-cultured rice calli was infected using the Agrobacterium-mediated transformation of rice (Oryza sativa L. cv. Taipei309) method, with an Agrobacterium clone carrying the CRISPR/Cas9 expression vector containing the Cas9 gene and a sgRNA targeting the OsAnn5 gene. For the expression vector corresponding to the targeted site in the fth exon of OsAnn5, 18 individual rice transgenic T 0 lines were obtained and were subjected to mutation detection by sequencing the PCR products harboring the sgRNA target sites. Only one mutant was identi ed and subjected to zygosity analysis by cloning PCR products into the T vector for DNA sequencing. The examination revealed the mutant was a homozygous biallelic mutant resulting from a 2-bp deletion (Fig. 5). For the expression vector corresponding to the targeted site in the third exon of OsAnn5, 34 individual rice transgenic T 0 lines were obtained. The sequence analysis revealed four types of Non-homologous end joining (NHEJ) mutations: +1 (1-bp insertion), -1 (1-bp deletion), -4 (4-bp deletion), and − 6 (6bp deletion) (Fig. 6). Out of the 3 mutants, two were monoallelic mutants and one of them was a heterozygous biallelic mutant. In view of the nding that the T 0 biallelic mutant progeny were all mutant, two T 1 mutant lines from the T 0 biallelic mutant (B1-KO-8 and B2-KO-21) were used for the identi cation of the cold tolerance phenotype. To examine the effect of the OsAnn5 gene knockout on cold tolerance, the four week old rice seedlings of the WT and KO lines were exposed to cold stress treatment (4 ~ 6 °C for 3 days), and then returned to the normal growth conditions to recover. After approximately 10 days in the greenhouse following cold treatment, the two T 1 KO lines re-grew 37.5% and 34.7%, respectively, while the survival ratio of the corresponding WT lines reached 80.5% and 78.2% respectively (Fig. 7). Under the same cold treatment conditions, the relative electric conductivity of leaves was measured. Results revealed that the relative electric conductivity levels in the two T 1 mutant lines were signi cantly increased after cold treatment in comparison to the WT, while they were similar in the two T 1 KO lines and WT non-stress conditions (Fig. 7). These results showed that the knockout of the OsAnn5 gene signi cantly decreased cold tolerance of rice at the seedling stage.

Potential Off-target Loci Analysis
In this study, potential off-target loci were analyzed using the CRISPR-GE software package (http://skl.scau.edu.cn/) (Xie et al., 2017). For the targeted site in the fth exon of OsAnn5, off-target loci prediction revealed three candidate sites which had 16-bp out of 20-bp identity and existed in exon regions of the targets Os07g0275475, Os11g0682300 and Os07g0598300 (Fig. 8). For the targeted site in the third exon of OsAnn5, there were two candidate sites which also had 16-bp out of 20-bp identity and existed in the exon region of the targeted Os03g0753500 and the intron of the targeted Os02g0654400 respectively (Fig. 9). The genomic sequence harboring the potential off-target site was ampli ed from WT rice and two T 0 biallelic mutants ((B1-KO-8 and B2-KO-21), and the PCR products were then sequenced. Overlapping signals and indels were not detected in our two T 0 biallelic mutants (Fig. S2). These results suggest that off-targeting did not take place in the evaluated candidate sites.

Discussions
With the increasing availability of genome sequencing, identi cation of rice annexin genes will continue to become easier. The role of rice annexins in responding to abiotic stress will also continue to be revealed. According to bioinformatics analyses, more than 20 putative cis-regulatory elements were identi ed in the OsAnn5 promoter region. Many of these are common promoter elements, including thirty-seven CAAT-box a (common cis-acting element in promoter and enhancer regions) and twenty-seven TATA-box (a core promoter element located around − 30 bp from the transcription onset). Some are unique to OsAnn5, including DRE core (a cis-acting element involved in CBF-mediated cold responsiveness) and MYB recognition sites. Additional cis-regulatory elements speci c to OsAnn5 were identi ed in the region between the start codon ATG and − 2082 bp, including plant hormone regulatory elements involved in methyl jasmonate, gibberellin and salicylic acid responsiveness, and several elements involved in light responsiveness. These ndings suggest that OsAnn5 may be regulated by both common and speci c transcription factors in rice.
We also demonstrated that OsAnn5-GFP fusion protein was mainly localized to the endoplasmic reticulum apparatus. Localization results were different in a recent analysis of the rice annexin gene OsANN1, whose subcellular localization was reported to be in the cytoplasm and cell periphery in the meristematic zone, and in the cell periphery in cells of the elongation zone (Qiao et al., 2015). Our ndings were also different to those related to the rice annexin gene OsANN3. OsANN3-GFP uorescence was observed in both the plasma membrane and cell periphery of rice root tip cells (Li et al., 2019). The variable subcellular localization patterns among different rice annexin genes may re ect the need for diverse functions. These differences may also be caused by other factors, such as phosphorylation of proteins and the internal and external environment of the cell and so on. Phosphorylation of AnxA2 protein leads to its translocation to the plasma membrane. It was suggested that phosphorylation processes might regulate annexin distribution between cellular compartments (Deora et al., 2004;Rescher et al., 2008). AtAnn1 was found to exist widely in the plasma membrane, mitochondria, cytoplasm, thylakoid and glyoxylate cycle (Laohavisit and Davies., 2011). The implication is that some plant annexins could be in different locations within the cell at the same time.
CRISPR/Cas9 technology has demonstrated enormous potential as an effective genome editing tool for basic and applied research in plants. In this study, in order to enhance mutation e ciency and ensure that mutations can result in the loss of the target gene function, we designed two target sites in different exons of OsAnn5, and mutant plantlets were successfully obtained in both sites. However, for one of the two target sites, only one mutant was detected in 18 individual rice transgenic T 0 lines, providing mutation e ciency of only 5.6%. Although we used the same CRISPR/Cas9-mediated expression vector backbones, this mutation e ciency is notably lower than the 31.6% mutation e ciency in our previous study (Shen et  Yang. 2013). For the study of gene function using CRISPR/Cas9-mediated gene editing, off-target activity can affect the nal phenotypic determination: once off-target activity takes place, it becomes di cult to determine whether phenotypic change is due to target gene knockout or off-target activity. Therefore, it is necessary to evaluate potential off-target loci when we carry out gene knockout using a CRISPR/Cas9-mediated approach. Off-target mutations caused by the CRISPR system can be minimized by choosing target sequences that have reduced numbers of off-targets. CRISPR-GE (http://skl.scau.edu.cn/) is a convenient and integrated toolkit by which we can expedite all experimental designs and analyses of mutations for CRISPR/Cas9 genome editing in plants, and it provides a set of powerful tools for prediction of off-target sites. In this study, we evaluated ve candidate off-targets, including two in the exons of OsAnn5 using the CRISPR-GE software package. We found no evidence for off-targeting phenomena in the candidate sites. These results support the reliability of our identi cation of the cold tolerance phenotypes in mutants.
Finally, our results demonstrated that based on the electrical conductivity and survival ratio tests, T 1 mutant lines from two T 0 biallelic mutants showed decreased cold tolerance compared with the Taipei309 WT variety. Additionally, overexpression of OsAnn5 improved tolerance to cold stress in rice. These results indicate that the rice annexin gene OsAnn5 is a positive regulator of cold stress tolerance at the seedling stage. OsAnn5 thus becomes only the second rice annexin gene reported to be involved in cold tolerance at the seedling stage, following previous reports of a similar role for the annexin gene OsAnn3. These results expand our understanding of the complex mechanisms of annexin response to cold stress in rice. Genetic engineering using annexin genes might offer a new and excellent platform to develop rice cold resistance breeding. Supplementary Information Table S1. Primers used in this study.    Overexpression of OsAnn5 resulted in increased cold tolerance. (A) Growth performance of OE and WT seedlings that were four weeks old, in 9 cm diameter glass petri dishes before and after stress (4~6 °C for 3 d). WT, wild type; OE, overexpression; BS, before stress; R-10d, recovery for 10 d after stress. The experiment was repeated three times. (B) Survival rate of OE and WT seedlings after stress. (C) Relative electrical conductivity of OE and WT seedling leaves before and after cold treatment. OE seedlings (OE-18, OE-23 and OE-24) showed higher survival rate lower relative electrical conductivity and better growth than WT seedlings. One asterisk indicates signi cant difference (P<0.05) in comparison with WT. Error bars represent the s.e.m.

Figure 8
Subcellular location of OsAnn5-GFP in rice protoplasts. A-F: The WT GFP and OsAnn5-GFP are separately transformed into rice protoplasts; G-J: Colocalization of OsAnn5-GFP with mcherry marker at endoplasmic reticulum(ER). Scale bars=10 μm; WT, wild type.    Potential off-targets at the Os07g0275475, Os11g0682300 and Os07g0598300 loci. Mismatches between potential off-target sites and the targeted region are indicated in red. The PAM sequences are underscored. Figure 18