The carbohydrate-binding module mediates mCherry protein anchoring on the cell wall in rice

The red fluorescent protein mCherry is widely used as a tagged protein in intracellular protein positioning and dynamic tracing due to its stable characteristics in color and monomer molecule. In this study, mCherry was used as a tag to explore the cell wall-directed binding of the carbohydrate-binding module (CBM), in order to intuitively demonstrate that the fusion protein anchored on the cell wall. Two constitutive expression vectors harboring CBM-mCherry fusion gene were constructed for rice transformation. The results of fluorescent signal detection showed that both the ubiquitin promoter and the CaMV35S promoter could drive the expression of mCherry fusion gene in seeds, leaves, and roots. The results of subcellular localization by different methods, such as cytological observation, immunofluorescence detection and protoplast observation, displayed that the fluorescence signals were concentrated on the cell walls of the root, stem, leaf, and vascular bundle sheath in rice. It indicated that the CBM could mediate the red fluorescent protein anchoring on the cell wall, and the cell wall may be a good subcellular structure for accumulation of exogenous proteins in future. The fusion gene SP-CBM11-mCherry was transformed into rice, and the subcellular localization of fusion protein indicated the red fluorescent protein mCherry anchored on the cell wall.


Introduction
The cell wall is one of the important features to distinguish plant cells from animal cells, and its main components are cellulose, hemicellulose and pectin (Gigli Bisceglia et al. 2020). It is also the primary barrier involved in immune or defense, and the chemical components and secondary metabolites of the cell wall played important roles in responding to pests and diseases (Le Gall et al. 2015). However, the major substance of the cell wall can be degraded by a consortium of microbial enzymes, such as cellulases and hemicelluloses, which contain catalytic domain (CD), carbohydrate-binding module (CBM) and linker (Boraston et al. 2004). CBM is a noncatalytic module found in those carbohydrate-degrading enzymes (Sidar et al. 2020), and a group of independent domains which can bind polysaccharides and bring the biocatalyst into intimate and proximity to its substrate, which increases the effective enzyme concentration on the polysaccharide surface and the rate of catalysis (Guillen et al. 2010). Till now, over 88 families of CBM are reported in the CAZy database (http:// www. cazy. org/). Based on the structural similarity and ligand binding function, CBM can be divided into three types. Type A CBM is a surface-bound style including family 1, 2a, 3a, 5, 10, and so on (Blake et al. 2006). The binding site of this type is distributed in or close to a plane, and selectively binds the insoluble crystalline substrate, such as crystalline cellulose or crystalline chitin (Gilbert et al. 2013). Type B CBM is a chain-bound type Communicated by Goetz Hensel. containing family 2b, 4, 6, 11, 15, 17, 20, 22, 27, 28, 29, 34, 36, etc. (Armenta et al. 2017). This type of CBM has a slit or cavity on the surface, which can bind to a wide variety of glycans, such as xylans, mannans, galactans and starch (Armenta et al. 2017). Type C CBM is a small molecule sugar-bound type (Boraston et al. 2004) including family 9, 13, 14, 18, etc., and their binding substrates are limited to monosaccharide, oligosaccharide, or polysaccharide terminal sugar groups . Therefore, the target and specificity for substrate recognition of carbohydrates make CBM a good domain structure for anchoring on the cell wall of plants.
Clostridium thermocellum is an anaerobic thermophilic cellulolytic bacterium. In this bacterium, bifunctional cellulase Lic26A-Cel5E contains the GH26 and GH5 catalytic modules that display β-1,4 and β-1,3 − 1,4-mixed linked endoglucanase activity, respectively (Carvalho et al. 2004), and the CtCBM11 structure of Lic26A-Cel5E in Clostridium thermocellum belongs to family 11 CBM of type B, which binds polysaccharides into their groove on the surface (Viegas et al. 2008). β-1,3 − 1,4-glucan is abundant in many cell wall polysaccharides of cereals and cereal endosperm, where are the main locations of accessible substrate for CBM11 targeting and adhering (Scheller and Ulvskov 2010). For instance, the CBM3 domain in BsCel5A-CBM3 replaced by CtCBM11 domain minimized the non-specific binding to the lignin component and enhanced the activity of β-glucanase (Fonseca Maldonado et al. 2017). It was also reported that the addition of CtCBM11 to the C-terminus of cellulase A could enhance protein stability under extreme pH conditions and increase the affinity and activity to insoluble polysaccharides (Cattaneo et al. 2018).
In addition to carbohydrates, there are small amounts of protein in the plant cell wall (Jamet et al. 2006). These proteins include extensins, glycine-rich proteins, prolinerich proteins, solanaceous lectins and arabinogalactan proteins (Somerville et al. 2004). Among these proteins, the extensins are a class of hydroxyproline-rich glycoproteins characterized by the signal peptide and repeating units of Serine-(Proline) n Showalter et al. 2010). Analysis of the extensins from maize (Stiefel et al. 1991), Nicotiana tabacum (De Loose et al. 1991), and bean (Wycoff et al. 1995) showed that these proteins contain signal peptide sequence at N terminus. It was demonstrated that the extensin signal peptide from Nicotiana plumbaginifolia could secrete heterologous proteins from protoplasts into the extracellular space (De Loose et al. 1991). When the signal peptide from the extensin gene of Arabidopsis thaliana fused with the CBM from nonhydrolyzed protein family 2 of strawberry, the fusion protein CBMFaEXP2 could be secreted and localized on the cell wall (Nardi et al. 2015). When the signal peptide of extensin was linked with the CtCBM11 and iron-binding peptide, the iron ion signal could be observed by x-ray fluorescence microscopy on the cell wall of the Arabidopsis epidermis (Yang et al. 2016). Therefore, the extensin signal peptide was useful for extracellular secretion of proteins.
Till now, most of the studies focused on the intracellular expression and accumulation of exogenous proteins in plants, and few of the studies involved the secretion of exogenous proteins to the cell wall. Based on the specific binding properties of CBM to cell wall polysaccharides and the extracellular secretion function of the extensin signal peptide (SP), we designed a fusion protein SP-CBM-mCherry to verify that the CBM could guide the directional accumulation of the red fluorescent protein mCherry on the cell wall, so as to provide references and technical supports for the extracellular secretion and directional accumulation of other important functional proteins on the cell wall.

Fusion protein design and vector construction
The sequence of SP and CtCBM11 come from Nicotiana plumbaginifolia (De Loose et al. 1991) and Clostridium thermocellum Lic26A-Cel5E (GenBank: P16218.1) (Carvalho et al. 2004), respectively. The sequence of mCherry protein was from the tension sensor gene in the synthetic construct YPet(short)-FL-mCherry (GenBank: MF685013.1) (Ringer et al. 2017). The sequence of the chymotrypsin recognition site (NPAAPFRNP, GenBank: GU327680.1) was used to connect CtCBM11 with mCherry. The fusion protein SP-CBM11-mCherry was assembled by successive linking the SP, CtCBM1, chymotrypsin recognition sequence and mCherry (Fig. 1a), and the target of fusion protein was predicted on line (https:// linux1. softb erry. com). Without altering the encoded amino acid sequence, the fusion gene SP-CBM11-mCherry was optimized according to the codon bias of rice, then the KpnI and SpeI restriction sites were added at the 5′ and 3′ end, respectively. The fusion gene was synthesized by BGI (Shenzhen, China) and cloned into the vector pUC57, and the recombinant vector was named as pUC57-CBM11mCherry (Fig. 1b). The 3D structure of the fusion protein was predicted online (https:// zhang group. org/I-TASSER/; Fig. 1c).

Transformation and identification of rice transformants
The rice japonica cultivar Nipponbare was used as the recipient and the strain EHA105 harboring pC3300-Epsps-UbiCBM11mCherry or pC3300-Epsps-35SCBM-11mCherry were used for Agrobacterium-mediated transformation, respectively (Hiei and Komari 2008). The rice genomic DNA was extracted by the CTAB method (Murray and Thompson 1980), and the primer pairs Epsps-gF/ Epsps-gR and mCherry-gF/mCherry-gR were used for detection of Epsps and SP-CBM11-mCherry, respectively ( Table 1). The PCR reaction volume was 20 µL including 1 µL genomic DNA template (30-50 ng/µL), 0.5 µL each of the forward and reverse primers (10 µmol/L), 10 µL of 2× M5 Taq HiFi PCR mix, and 8 µL ddH 2 O. The PCR procedure was conducted as follows: initial denaturation at 95 °C for 3 min; 35 cycles at 95 °C for 25 s, 58 °C for 25 s, 72 °C for 30 s; and final extension at 72 °C for 10 min.

Total RNA extraction and RT-PCR
Total RNA was isolated from transformants and wild type using the AG RNAex Pro Reagent (AG21101, Accurate Biology, Changsha, China) according to the manufacturer's instruction. The template cDNA was synthesized using 1 µg total RNA via 5× All-In-One RT Master Mix Kit (ABM, Catalog#G486). RT-PCR was performed using specific primers for detecting the expression of Epsps and mCherry (Table 1). The RT-PCR procedure was as follows: initial denaturation at 94 °C for 3 min; 30 cycles at 94 °C for 30 s, 60 °C for 30 s, 72 °C for 30 s; with a final extension at 72 °C for 4 min. The gene actin7 was used as the internal control (Fu and Liu 2020).

Glyphosate tolerance assay
About sixteen sterilized seeds from transformants and wild type were inoculated on solidified MS medium containing 340 mg/L glyphosate, and the wild type seeds without glyphosate stress were used as control. All of them were cultivated in a growth chamber at 28 °C and with a 12 h/12 h light/dark photoperiod. After cultivation for 10 days, the plant status was photographed, as well as the root length, shoot length and fresh mass were measured.

Fluorescence analysis
The seeds from transformants and wild type were observed under white and red filters by plant vivo imaging system (FUSION FX7 EDGE SPECTRA). The roots and leaves from transformants and wild type were free-hand sectioned using a razor blade and then the sections were observed under confocal laser microscope (LSM880, CarlZeiss) at excitation and emission wavelengths (587 nm/610 nm) of mCherry.

Subcellular localization of fusion protein
The calli of transformants and control were placed on glass slides and scanned by confocal laser microscope (LSM880, CarlZeiss) at 587 nm excitation wavelength and 610 nm emission wavelength.
The leaves from 30-day-old seedlings were used for paraffin embedding. The paraffin-embedded sections were incubated at room temperature in a 1:200 dilution of Anti-mCherry-Tag Mouse Monoclonal Antibody (9D3) (Abbkine, A02080) for 1 h, then the sections were washed three times for 5 min with 1× TBST buffer and incubated at room temperature in dark for 50 min in a 1:200 dilution of DyLight488 Goat Anti-Mouse IgG (Abbkine, A23210) as the second antibody. The sections were stained with 1 mg/ mL DAPI (4′,6-Diamidine-2′-phenylindole) for 10 min (Tan

Protoplast observation
Etiolated seedlings that grew at 28 °C for about 12 days in dark were used for isolation of protoplasts. Before isolation of protoplasts, the leaf sheaths were used for detection of the red fluorescence of mCherry, and the Calcofluor White was applied as a specific dye of cell walls. The Calcofluor White was excited at 405 nm, and the emission was detected at 425-475 nm. The protoplasts were isolated according to the reference (He et al. 2016) and observed at excitation and emission wavelengths (587 nm/610 nm) for detection of mCherry with laser microscope (LSM880, CarlZeiss).

Construction of vectors with fusion gene SP-CBM11-mCherry
The CBM11 and mCherry were linked by the chymotrypsin recognition site, which facilitated both site-directed digestion in vitro and inter-domain flexibility (Fig. 1a). The fusion gene SP-CBM11-mCherry was cloned in vector pUC57-CBM11mCherry (Fig. 1b) and verified by digestions of KpnI and SpeI restriction endonuclease (Fig. 1f).
The structure prediction of fusion protein displayed that the spatial structure of CBM11 and mCherry was independent of each other without cross-folding, and the two domains formed their active spatial conformations, respectively (Fig. 1c). The vectors pC3300-Epsps-UbiCBM11mCherry (Fig. 1d) and pC3300-Epsps-35SCBM11mCherry (Fig. 1e) were constructed based on pC3300-Epsps-UbiCry1Ca and pC3300-Epsps-35SCry1Ca respectively, and the construction of them was verified to be successful by digestion of restriction enzymes (Fig. 1g).

Identification of rice transformants
The regenerated plantlets were produced after herbicide selection and differentiation (Fig. 2a -c), and the seeds with red fluorescence were harvested after the regenerated plants grew in pots for about three months (Fig. 2d). The integrations of genes Epsps and SP-CBM11-mCherry in regenerated plants from transformation of two expression vectors were verified by primers Epsps-gF/Epsps-gR and mCherry-gF/mCherry-gR (Table 1), respectively ( Fig. 2e and f). The results showed that the regenerated plants from transformation of pC3300-Epsps-UbiCBM11mCherry can amplified 1363 bp and 704 bp target fragments. Most of the regenerated plants from transformation of pC3300-Epsps-35SCBM11mCherry could detect the target fragments of 1363 bp and 704 bp. The offspring were selected in successive generations to seek stable transformants by herbicide stress and PCR selection (Fig. 2g, h), and the expression of target genes was also identified by RT-PCR (Fig. 2j). The above results demonstrated that the exogenous target genes Epsps and mCherry were successfully transferred into the rice genome.

Selection by red fluorescence and glyphosate resistance
Although some regenerated plants harbored the mCherry gene, their seeds did not display red fluorescence. Thus, phenotyping was necessary. The transformants bearing seeds with intensively red fluorescence were chosen for further analysis. Finally, three lines named as OE-1, OE-2 and OE-3 from transformation of pC3300-Epsps-UbiCBM11mCherry, and three lines named as OE-4, OE-5 and OE-6 from transformation of pC3300-Epsps-35SCBM11mCherry were selected out ( Fig. 3a and b). Both the seeds (Fig. 3a) and endosperm (Fig. 3b) of them displayed a cherry pink color under natural light, and both the unhusked rice and brown rice of them showed dark red fluorescence under red light (Fig. 3c), that indicated the mCherry protein was a useful reporter in genetic transformation. The herbicide resistance was verified on a solid MS medium containing 340 mg/L glyphosate. The results displayed that the transformants normally germinated and grew under glyphosate stress, while the wild type did not germinate under same stress (Fig. 4a), that was in accord with the results of fluorescence screening (Fig. 4b). The root length (Fig. 4c), shoot length (Fig. 4d), and fresh mass (Fig. 4e) of transformants were not significantly different to those of wild type without herbicide stress respectively, whereas the growth of wild type with herbicide stress was severely inhibited (Fig. 4a). The above results indicated both the glyphosate tolerance and red fluorescence could be used for transformant selection.

Cytological observation of red fluorescence in leaves and roots
The transformant OE-1 was used to confirm the expression of mCherry gene by observation of red fluorescence on its leaf (Fig. 5a), root elongation zone (Fig. 5b) and root tip (Fig. 5c). In leaf section, the red fluorescence existed mainly in the cell wall and vascular bundles, as well as scattered lightly in cytoplasm of transformant, whereas the wild type did not show red fluorescence (Fig. 5a). In section of and RT-PCR identification of genes SP-CBM11-mCherry and Epsps in T 3 generation. 1, 2, 3 and 4: transformants from transformation of pC3300-Epsps-UbiCBM11mCherry; 5, 6, 7 and 8: transformants from transformation of pC3300-Epsps-35SCBM11mCherry root elongation zone, the red fluorescence distributed along the profile of cell wall (Fig. 5b). In root tip section, the red fluorescence should be distributed on the cell wall, but that looked like filling in the whole cell due to the low resolution or over exposure (Fig. 5c).

Observation of callus fluorescence and immunofluorescence of leaf section
At in vitro selection and differentiation stage, the transformed calli showed a visible red color ( Fig. 2a and b). Microscopic observations indicated that the red fluorescence of transformed calli existed in the cell wall of parenchyma cells, whereas the calli of wild type did not show red fluorescence (Fig. 6a). The immunofluorescent observations indicated that the fluorescence signals localized at the vascular bundle sheath cell wall and phloem, where were rich in polysaccharides that overlayed the cell wall structure (Fig. 6b).

Observation of protoplast fluorescence
The protoplasts from rice stable transformants harboring fusion gene were isolated for red fluorescence observation.
Before protoplast isolation, the cells were outlined by cell walls with red fluorescence (Fig. 7a). After the cell walls were removed, there was no red fluorescence surrounding the protoplasts, and the protoplasts became spherosome without restriction of cell walls (Fig. 7b). The above results indicated that the fusion protein with red fluorescence was secreted and anchored on the cell wall.

Discussion
The signal peptide is crucial for the localization and stability of foreign proteins. By use of transit peptide of chloroplast, the subcellular localization of proteins Cry1Ac and Cry2A provided an effective alternative way for insect management in cotton (Muzaffar et al. 2015). The modified Cp4-EPSPS with an N-terminal chloroplast targeting peptide increased glyphosate tolerance (Karthik et al. 2020). By using the retention signal peptide of endoplasmic reticulum, the CpTI protein accumulated more in transgenic tobacco (Deng et al. 2003). Using endosperm-specific promoter Gt13a and a signal peptide targeting the OsrhGM-CSF to the protein bodies, Ning et al. achieved the high expression level of OsrhGM-CSF protein in rice endosperm (Ning et al. 2008). Wang et al. also reported that a novel phytase-lactoferricin fusion gene driven by endosperm-specific promoter Gt13aP with its signal peptide sequence Gt13aSP could improve the phosphorus availability and antibacterial activity of rice seeds (Wang et al. 2017). However, the loading of exogenous proteins inside the cell is limited, the cell wall may be a good subcellular structure for directed accumulation of exogenous proteins. Actually, the cell wall is also one of the sites for protein accumulation (Calderan-Rodrigues et al. 2019). Extensins are hydroxyproline-rich glycoproteins and the most abundant structural proteins in the cell wall . Their signal peptides at the N-terminus were responsible for the secretion of extensin into the extracellular matrix (Showalter et al. 2010). For example, the β-expansins GmEXPB2 (Guo et al. 2011) andOsEXPB2 (Zou et al. 2015) in root were characterized as secretory proteins located on the cell wall, and both of them were presumed to comprise signal peptides at N-terminus. By fusing a signal peptide from Arabidopsis expansion AtEXP8 at the N-terminal of CBM, the fusion protein CBMFaEXP2 was targeted to the cell wall (Nardi et al. 2015). In this study, the signal peptide from N. plumbuginifolia (De Loose et al. 1991) was used for secretion of red fluorescent protein mCherry to the extracellular, that was confirmed by red fluorescence detection in rice endosperm (Fig. 3b) and cell wall (Figs. 5 and 6). Furthermore, when the plant cell wall was degraded by cellulase and pectinase, there was no red fluorescence surrounding the protoplasts (Fig. 7b), that proved on the other hand the fusion protein was secreted to the cell walls.
When the foreign protein was shunted into the secretion pathway, it was randomly diffused into the extracellular matrix. Therefore, the adherence of foreign protein is also crucial for its accumulation. The CBMs are independently structural domains that can recognize polysaccharides, such as cellulose and starch (Armenta et al. 2017). The position of CBM located at N-terminal or C-terminal of the fusion protein does not affect the function of CBM (Foumani et al. transformation of pC3300-Epsps-UbiCBM11mCherry; OE-4, OE-5 and OE-6: transformants from transformation of pC3300-Epsps-35SCBM11mCherry; CK-0: wild type without herbicide stress; CK-T: wild type with 340 mg/L glyphosate stress; ns: no significant difference; ***: significant difference at the 0.001 level using oneway ANOVA 2015). In this study, the CtCBM11 was used to anchor the red fluorescent protein to cellulose in the cell walls, that was derived from Clostridium thermocellum and could specifically bind the glucans linked by 1,4-β and 1,3 − 1,4-β glycosidic bond (Carvalho et al. 2004;Ribeiro et al. 2020;Viegas et al. 2013). The structure prediction of CtCBM11 demonstrated that it consisted of a β-sandwich that formed a concave of the substrate-binding cleft (Fig. 1c), and the fluorescence analysis showed that the fusion protein accumulated in rice endosperm that was rich in starch ( Fig. 3a ~ 3b) or cell walls that were abundant in carbohydrates such as cellulose (Fig. 5). Moreover, callus fluorescence and leaf sheath immunofluorescence demonstrated that the fusion protein localized on the cell wall and leaf vascular bundle sheath, respectively (Fig. 6). In summary, our study demonstrated that the extensin signal peptide guided the secretion of the fusion protein to the extracellular matrix, and the CBM domain anchored the fusion protein on the cell walls.
The promoter is like a switch in regulating the activity of downstream genes, playing an important role in gene expression. Application of constitutive promoters endows the plants with constant and high expression level in whole tissues (Feike et al. 2019). The CaMV35S promoter was frequently implemented in dicot plants such as soybean (Yang et al. 2021) and cotton (Sufyan Tahir et al. 2021), whereas the maize ubiquitin promoter was mostly used in  (Achary et al. 2020) and maize (Peterson et al. 2021). In this study, the fusion gene SP-CBM11-mCherry was driven by ubiquitin promoter and CaMV35S promoter, respectively. Both of the promoters could get a satisfactory expression level of the fusion protein based on the observation of naked eyes ( Fig. 3a and b). The application of red fluorescent protein in tracking gene expression in plant could decline the interference of chlorophyll in green tissue (Fig. 5a) and be convenient for visual selection ( Fig. 2a-d, Fig. 3a and b), that would save labor and speed up the selection process of transformants. Contrary to the positive selection by visible red protein, the expression of herbicide resistance gene Epsps was convenient for killing wild-type cells by herbicide at an early stage when the positive selection was invalid or inefficient even though the cells showed the visible red color. At the seedling stage, the herbicide resistance of genetically modified plants also facilitated the selection on MS medium containing glyphosate (Fig. 4a and c-e) and in the fields. In terms of efficiency, we thought the negative selection like herbicide selection was irreplaceable.
Author contributions HL designed, did the experiments and wrote the manuscript; GX conceived the project, secured the funding and revised the manuscript; DL, WL, LJ and YJ contributed to the planting management in the field. All authors have read and agreed to the published version of the manuscript.

Declarations
Conflict of interest All authors declare no conflicts of interests. The manuscript that a non-peer reviewed preprint of this work has been posted in ResearchGate.