Development of a split fluorescent protein-based RNA live-cell imaging system to visualize mRNA distribution in plants

Background RNA live-cell imaging systems have been used to visualize subcellular mRNA distribution in living cells. The RNA-binding protein (RBP)-based RNA imaging system exploits specific RBP and the corresponding RNA recognition sequences to indirectly label mRNAs. Co-expression of fluorescent protein-fused RBP and target mRNA conjugated with corresponding RNA recognition sequences allows for visualizing mRNAs by confocal microscopy. To minimize the background fluorescence in the cytosol, the nuclear localization sequence has been used to sequester the RBP not bound to mRNA in the nucleus. However, strong fluorescence in the nucleus may limit the visualization of nucleus-localized RNA and sometimes may interfere in detecting fluorescence signals in the cytosol, especially in cells with low signal-to-noise ratio. Results We eliminated the background fluorescence in the nucleus by using the split fluorescent protein-based approach. We fused two different RBPs with the N- or C-terminus of split fluorescent proteins (FPs). Co-expression of RBPs with the target mRNA conjugated with the corresponding RNA recognition sequences can bring split FPs together to reconstitute functional FPs for visualizing target mRNAs. We optimized the system with minimal background fluorescence and used the imaging system to visualize mRNAs in living plant cells. Conclusions We established a background-free RNA live-cell imaging system that provides a platform to visualize subcellular mRNA distribution in living plant cells.


Background
Most cells transport mRNAs to specific subcellular compartments for localized RNA translation. Instead of transporting a quantity of proteins, the transport of mRNA provides an efficient strategy to distribute proteins to specific subcellular compartments. In eukaryotic cells, asymmetric mRNA localization is a conserved mechanism to create translation hotspots for regulating spatial gene expression in response to environmental cues [1]. The intracellular mRNA asymmetric distribution in plants can be extended to the intercellular level in which specific plant mobile mRNAs are transcribed in local tissues and undergo long-distance movement via phloem to exert non-cell autonomous functions, including regulation of leaf development [2], floral transition [3,4], and potato tuber formation [5]. Thus, understanding the mRNA subcellular distribution may provide information to unravel RNA logistics and post-transcriptional gene regulation.
In living plant cells, direct visualizing of mRNAs by microscopy is challenging. Many approaches have been developed that allow visualizing mRNAs in living cells. Direct labeling of mRNAs can be achieved by incorporating in vitro-synthesized mRNA with fluorescein-labeled nucleotides or introducing molecular beacons into cells to hybridize with the target mRNA [6,7]. However, these in vitro RNA labeling systems require invasive injection of labeled RNA into plant cells such as microinjection or particle bombardment. In addition, the in vitro-synthesized RNAs may behave differently from the in vivo-transcribed RNAs because the RNA synthesized in vitro may bypass the RNA modification occurring in the nucleus, which may change the properties of RNA.

Plant Methods
Alternatively, RNA can be indirectly visualized by an RNA aptamer-based approach. In this system, the target mRNA is genetically modified by tagging with an RNA aptamer recognized by RNA-binding proteins (RBPs). Co-expression of fluorescent proteins (FPs) conjugated with RBPs with aptamer-tagged mRNA allows for indirect labelling of mRNA in living cells. MCP, a coat protein derived from MS2 bacteriophage with RNA-binding activity, and the stem loop structures recognized by MCP have been widely used to monitor mRNA localization in yeast, fruit fly, mammals and plants [8][9][10][11][12][13]. To reduce the background GFP fluorescence detected in the cytosol, the nuclear localization sequence (NLS) has been used to sequester the MCP-GFP not bound to RNA in the nucleus [8,13]. In addition, the signal-to-noise ratio of RNA detected in the cytosol can be greatly improved by the plant-specific NLS tightly restricting MCP in the nucleus [13].
Although the MCP-based approach has been successfully used to visualize the intracellular distribution of mobile and non-mobile mRNA in plants [13], the background GFP fluorescence derived from nucleus-localized MCP-GFP may interfere with detecting mRNA located close to the nucleus (Fig. 1A). In addition, cells that undergo active cell division are smaller and the nucleus is relatively large [14], so the high nucleus-to-cytoplasmic ratio may disturb the detection of RNA localization due to strong background GFP fluorescence in the nucleus (Fig. 1B).
In this study, we improved the RNA live-cell imaging system by eliminating the background GFP fluorescent in the nucleus. The application of this split FPs-based system detected differential subcellular distribution between mobile and non-mobile mRNAs. In addition, the split FPs-based system greatly improved the signal-to-noise ratio in the nucleus, which confers detection of mRNAs in the nucleus or at juxtanuclear region. Thus, our system broadens the application of RNA live-cell imaging system in plants.  [13]. The background fluorescence derived from nucleus-localized MCP-GFP interferes with the detection of potential FT mRNA foci (arrows) in the nucleus. Nucleus boundary is outlined with white dots. Bar = 2 μm. Inset, in cells with high accumulation level of MCP-GFP in the nucleus, high nucleus-to-cytoplasmic ratio of GFP fluorescence disturbs detection of FT mRNA foci (arrowheads) proximal to nucleus. Bar = 5 μm. B Roots of Arabidopsis transformant harboring constitutively expressed MCP FD -GFP. Note that in cells with high nucleus-to-cytoplasmic ratio, the background GFP fluorescence may interfere in mRNA localization in the cytosol. Scale bar = 20 mm

Development of split FP-based RNA live-cell imaging system
To develop the mRNA live-cell imaging system without background GFP fluorescence in the nucleus, we exploited the bimolecular fluorescence complementation system to reduce background GFP fluorescence derived from MCP-GFP not bound to mRNA. In this system, two different RBPs were fused with the N-or C-terminus of split FPs to reduce the fluorescence produced from RBP-FPs not bound to mRNA. The target mRNA was flanked with the corresponding RNA recognition sequences of the RBPs. The interaction between RBPs and RNA recognition sequences may bring two split FPs together to reconstitute a mature FP ( Fig. 2A).
We selected two coat proteins (CPs) of MS2 and PP7 bacteriophages, MCP and PCP, which specifically bind to an RNA hairpin structure of MBS and PBS, respectively [15]. MCP and PCP were conjugated with the N-terminus fragment (FP N , amino acids 1-173 from Venus) or C-terminus fragment (FP C , amino acids 156-239 from SCFP3A) of the split FP protein [16]. To identify the optimized fusion proteins for the system, we generated different MCP/PCP-FP fusion proteins by flanking FP N or FP C with MCP or PCP (Fig. 2B). In addition, the NLS derived from Arabidopsis FD [13], which encodes a bZIP transcription factor [17], was used to locate MCP or PCP to the nucleus (Fig. 2B).
To generate the target mRNAs that can be recognized by MCP and PCP, the DNA fragments containing 12   we transiently expressed different MCP-and PCP-split FPs in leaves of 3-weak-old Nicotiana benthamiana by agro-infiltration and examined GFP fluorescence under a confocal laser scanning microscope at 3 days after infiltration. When NLS-containing M NLS -FP N was coexpressed with P-FP C or FP C -P, weak GFP fluorescence was detected in both the cytosol and nucleus (Fig. 3A-D), which suggests that the separation of MCP and PCP into different subcellular compartments may not significantly reduce the background GFP fluorescence. Consistently, when NLS-containing P NLS -FP C was co-expressed with FP N -M, the background GFP fluorescence was again detected in the cytosol and nucleus (Fig. 4A, D, G). However, when P NLS -FP C was co-expressed with M-FP N or NLS-containing M NLS -FP N , the background GFP fluorescence that formed a speckle-like spot was detected only in the nucleus (Fig. 4B, C, E, F, H, I), which suggests that the topology of fusion proteins rather than subcellular localization may contribute to the background GFP fluorescence.
We next examined background GFP fluorescence in cells expressing MCP-FP and PCP-FP without NLS. When M-FP N was co-expressed with P-FP C , we consistently obtained the images without detectable background GFP fluorescence (Fig. 5A and E). However, the other 3 combinations of MCP-FP and PCP-FP conferred weak ( Fig. 5B and F) or strong GFP fluorescence (Fig. 5C, D, and G, H). We summarize these results in Table 1. Taken together, our results reveal that the combination of MCP-FP N and PCP-FP C produced undetectable background GFP fluorescence. In addition, the combination of P NLS -FP C with M-FP N or M NLS -FP N produced specklelike background GFP fluorescence in the nucleus, which  may also provide a useful positive control for the presence of the infiltrated constructs.

Detection of mRNA localization in background-free split FP-based imaging system
To verify the application of the split-FP system for localization of mRNAs, we co-infiltrated M-FP N and P-FP C with FT HSL12 or RFP HSL12 , two mRNAs with differential subcellular distribution pattern [13] in tobacco leaves. At 3 days after infiltration, no fluorescence was detected in control cells co-expressing M-FP N and P-FP C without mRNA (Fig. 6A, E). However, GFP fluorescence was detected in the cytosol of cells co-expressing M-FP N and P-FP C with RFP HSL12 or FT HSL12 mRNA. Of note, the fluorescence in the cytosol showed distinct distribution patterns in cells expressing RFP HSL12 or FT HSL12 mRNA: in cells expressing RFP HSL12 , the fluorescence was evenly distributed at the cell periphery (Fig. 6B, F), whereas with FT HSL12 , the fluorescence showed a punctate distribution ( Fig. 6C, G). These subcellular distribution patterns were reminiscent of previous MS2-based mRNA live-cell imaging [13]. However, in the split FP-based imaging system, the GFP fluorescence spots were detected inside or in the boundary of the nucleus (Fig. 6D-H), which were not detected in the previous MS2-based system [13]. These results suggest the broader use of a split FP-based imaging system for visualizing mRNA distribution. We further verified the application of the split FPbased imaging system with speckle-like background GFP fluorescence for mRNA visualization. In control cells co-expressing M NLS -FP N and P NLS -FP C , speckle-like GFP fluorescence was detected in the nucleus (Fig. 7A and C), whereas in cells co-expressing M NLS -FP N and P NLS -FP C with FT HSL12 mRNA, the GFP fluorescence was detected in the cytosol with a punctate distribution pattern ( Fig. 7B and D). Of note, GFP fluorescence spots were detected in the nucleus (Fig. 7D), suggesting that

Conclusions
We have established a split FP-based RNA live-cell imaging system for visualizing subcellular mRNA distribution in living plant cells. This system eliminates the background GFP fluorescence in the nucleus. We applied this system to monitor intracellular transport of mobile and non-mobile mRNA. In addition, this system enables the detection of mRNAs near the nucleus. Our method broadens the application of RBP-based RNA imaging system in living plant cells.

Plant materials and growth conditions
Nicotiana benthamiana plants were grown in a growth chamber under 27 °C, 16-h/8-h day/night cycles and light intensity 100 μmol m −2 s −1 . The 3-week-old plants were used for agro-infiltration assay.

Plasmid construction and Arabidopsis transformation
The DNA fragments of MCP and PCP were PCR-amplified with the following primers: MCP-For The DNA fragments of MCP and PCP were confirmed by sequencing and cloned into split FP-containing binary vectors pVYNE, pVYNE(R), pSCYCE, or pSCYCE(R) [16] to generate M-FP N , FP N -M, P-FP C and FP C -P, respectively. The C-terminal fragment (211-240 a.a.) of Arabidopsis FD, which contains the nuclear localization sequence (NLS) [13], was inserted in M-FP N or P-FP C to generate M NLS -FP N and P NLS -FP C , respectively.
For the 12 copies of the MBS-PBS hybrid stem loop sequence (HSL12), the Pcr4-12X-MBS-PBS [15] DNA was digested with BamHI and SpeI, followed by the Klenow fill-in reaction. The 1402-bp HSL12 fragment was gel-eluted and ligated with the binary vector pCAMBIA1390 containing the 35S-FT or 35S-RFP to form the p1390-35S-FT HSL12 or p1390-35S-RFP HSL12 construct, respectively.

Agro-infiltration of N. benthamiana
Agrobacterium tumefaciens strain AGL1 carrying individual constructs was cultured in 20 mL LB broth (10 g/L tryptone, 5 g/L yeast extract, 10 g/L NaCl) containing 50 μg mL −1 kanamycin, 10 mM MES, pH 5.7 and 20 μM acetosyringone at 28 °C with 220 rpm shaking for 2 days. Bacteria were pelleted by centrifugation and resuspended in infiltration medium (10 mM MgCl2, 10 mM MES, pH 5.7, and 200 μM acetosyringone) to OD600 = 1.0. The bacterial solution was placed at room temperature for at least 1 h. For agro-infiltration, the bacteria solution was infiltrated into the abaxial side of the 4 th or 5 th leaves of 3-week-old N. benthamiana by using a needle-removed 5-mL syringe. For co-infiltration with 2 or 3 different constructs, different bacterial solution or infiltration media were mixed at a 1:1 volume ratio and infiltrated into leaves. The infiltrated plants were cultured in growth chambers for 3 days, then infiltrated leaves were excised for confocal laser scanning microscopy.

Confocal laser scanning microscopy
GFP fluorescence was observed by confocal laser scanning microscopy (LSM880, Carl Zeiss) with the argon laser set to 488/500-530 for excitation/emission.