The plasmodesmata-associated β-1,3-glucanase gene GhPdBG regulates fiber development in cotton

Yijie Fn, Shungshung Lin, Ynhui Lyu, Hihong Shng, Youlu Yun, Zhengmin Tng,Chengzhi Jio, Aiyun Chen, Piyi Xing, Li Zhng, Yuxio Sun, Hixi Guo, Tongtong Li,Zhonghi Ren, Fnchng Zeng,*

a State Key Laboratory of Crop Biology, College of Agronomy, Shandong Agricultural University, Tai’an 271018, Shandong, China

b State Key Laboratory of Cotton Biology, Key Laboratory of Biological and Genetic Breeding of Cotton, The Ministry of Agriculture, Institute of Cotton Research, Chinese Academy of Agricultural Sciences, Anyang 455000, Henan, China

c Zhengzhou Research Base, State Key Laboratory of Cotton Biology, Zhengzhou University, Zhengzhou 450001, Henan, China

d State Key Laboratory of Crop Biology, College of Horticulture Science and Engineering, Shandong Agricultural University, Tai’an 271018, Shandong, China

Keywords:Fiber/trichome β-1,3-glucanase Functional analysis Evolutional variation

ABSTRACT Trichomes are specialized structures that originate from epidermal cells of organs in higher plants.The cotton fiber is a unique single-celled trichome that elongates from the seed coat epidermis.Cotton(Gossypium hirsutum) fibers and trichomes are models for cell differentiation.In an attempt to elucidate the intercellular factors that regulate fiber and trichome cell development,we identified a plasmodesmal β-1,3-glucanase gene(designated GhPdBG)controlling the opening and closing of plasmodesmata in cotton fibers.Structural and evolutionary analysis showed haplotypic variation in the promoter region of the GhPdBG gene among 352 cotton accessions, but high conservation in the coding region.GhPdBG was expressed predominantly in cotton fibers and localized to plasmodesmata (PD).Expression patterns of PdBG that corresponded to PD permeability were apparent during fiber development in G.hirsutum and G.barbadense.The PdBG-mediated opening-closure of PD appears to be involved in fiber development and may account for the contrasting fiber traits of these two species.Ectopic expression of GhPdBG revealed that it functions in regulating fiber and trichome length and/or density by modulating plasmodesmatal permeability.This finding suggests that plasmodesmal targeting of GhPdBG,as a switch of intercellular channels, regulates single-celled fiber and trichome development in cotton.

1.Introduction

Trichomes are specialized epidermal cells found in most terrestrial plants and can be derived from all kinds of organs, including leaves,stems,seeds,and roots.There are many kinds of trichomes:multicellular and unicellular,unbranched and branched,and nonglandular and glandular secreting trichomes (GSTs) [1].Plant trichomes function in biotic and abiotic stress tolerance, including by preventing excessive transpiration, increasing freezing tolerance, regulating surface temperature, and resisting insect predation [2–4].GSTs can also secrete secondary metabolites [5],protect plants from insects and pathogens, and attract pollinators[6,7].

Cotton fibers are single cells extending from the epidermis of ovule outer integuments,and are the longest single cell type found in any higher plant [8].Owing to their unicellularity and linear form, they are an excellent model for research on cell differentiation and elongation in higher plants[9,10].Cotton fiber cell differentiation and development can be divided into four overlapping stages: initiation,rapid elongation,secondary cell wall thickening,and maturation-associated dehydration [11,12].Around 30% of epidermal cells in the seed coat of tetraploid cotton grow into specialized fibers[9].Traits associated with fiber quantity and quality differ among cotton germplasm resources such as G.hirsutum and G.barbadense, with the latter having the longer fibers.Elucidating the cellular and molecular basis of fiber elongation would shed light on the basic cell biology mechanism of fiber development and the determinants of fiber yield and quality.

The plant cell structure that has been proposed to be associated with fiber and trichome cell elongation is the plasmodesmata(PD)which are present at the cotton fiber’s base located underneath the seed coat [13,14].PD are connections between plant cells that serve as channels for cell-to-cell communication and function in cell fate determination and cell development coordination[13,15–17].A PD gate in a given trichome or fiber cell can be reversed during development [8].The gene expression that controls PD opening and closing appears to be associated with fiber and trichome cell elongation [14].

Modifications in callose deposition and PD structure affect the regulation of intercellular communication by PD [13].Callose, a polymer of 1,3-glucans, is deposited at and surrounds PD [18,19].

Callose deposition at PD decreases the trans-PD cytoplasmic channel’s size exclusion limit, reducing the symplast’s permeability [20–23].Several cellular components that regulate PD permeability have been characterized.The balance between β-1,3-glucanase (degrading) activity and callose synthase appears to be the critical factor controlling callose turnover at the PD for intercellular communication.β-1,3-glucanase gene expression degrades callose,and the content of callose regulates the PD channel switch and affects intercellular substance transport and signaling[24–27].

Many β-1,3-glucanase family gene products exert various functions in plants [20,24,28,29].Only specific members of the family participate in regulating PD-associated callose [30,31].The

AtPdBG1, AtPdBG2, AtPdBG3, AtPdBG4, and AtBG_ppap genes in Arabidopsis [23,32,33], and the GhPdBG3 in cotton [34] are located in the PD.A phylogenetic tree of the β-1,3-glucanase gene family,which includes the above known genes and the PdBG of our study,has been constructed [22,35].

Elucidation of cotton fiber and trichome cell development depends largely on knowledge of the function of intercellular channels mediating in development.The switch that controls trichome PD is still unknown [34].This study identified a regulatory gene that controls these intercellular channels and characterizes gene functions associated with cotton trichome development.

2.Materials and methods

2.1.Plant material

The cotton wild-type (WT) species G.hirsutum (TM-1) and G.barbadense(Hai 7124) were cultivated at the experimental station of Shandong Agriculture University.Flowers were labeled on the day of flowering to record ovule age.Cotton ovules were sampled at various development stages from 0 days post-anthesis (DPA) to 25 DPA.Ovules were excised from the ovaries and intact fibers were removed from the ovules.TM-1 was used as cotton genetic transformation material.Nicotiana benthamiana was used for subcellular localization and N.tabacum for subcellular colocalization.

2.2.Variation analysis of the PdBG gene coding sequence (CDS) and promoter in Gossypium germplasm

A haplotype analysis of the PdBG coding region and 2 kb upstream promoter region was based on 352 worldwide Gossypium germplasm accessions, the germplasms are from Wang et al.[36].These accessions were from seven ecoregions: cotton from North America (the North American group), from Mexico (the Mexico group), and from groups of cotton accessions in China.The China-associated groups were obtained from the main cottongrowing regions of China, including the north China region(NCR),which is the specific early maturation region,the northwest China region (NWCR), the Yellow River region (YRR), the Yangtze River region(YtRR),and Chinese cotton varieties except those from NCR, NWCR, YRR, and YtRR (the China group).For evolutionary phylogenetic tree construction and sequence alignment of PdBG,the amino acid sequences and promoter sequences of GhPdBG and GbPdBG genes were retrieved from the respective Dt subgroups of G.hirsutum and G.barbadense in the CottonGen database(https://www.cottongen.org/) [37].Using DNAMAN software version 6.0(LynnonBiosoft,Quebec,Canada)with default parameters,sequence alignments were compared.

2.3.RNA extraction and qRT-PCR

Total RNA of various cotton tissues was extracted with a Total Plant RNA Extraction Kit (Tiangen, Beijing, China, DP441).Easy-Script One-Step gDNA Removal and cDNA Synthesis SuperMix(Transgen, Beijing, China, AE311) was used for complementary DNA (cDNA) synthesis.UltraSYBR Mixture (Low ROX) (CW2601,CWBIO, Jiangsu, China) was used for quantitative real-time PCR(qRT-PCR).Amplification was performed following this protocol:95 °C for 10 min, 40 cycles of 95 °C 15 s, and 60 °C for 30 s, and 72 °C 30 s.All reactions were performed in triplicate using three biological replicates for each RNA sample.GhUB7 (ubiquitin7) of G.hirsutum was used as an internal reference gene to standardize gene quantitative expression analysis.qRT-PCR primers are listed in Table S1.

2.4.Subcellular localization and colocalization

pGWB4 and pGWB454 were used to construct green fluorescent protein (GFP) and red fluorescent protein (RFP) fusion proteins,respectively.Without a stop codon,the full-length cDNA of GhPdBG was inserted into pGWB4.The constructs were transferred into N.benthamiana leaves by agroinfiltration and fluorescence signal was detected 72 h after injection [38].For subcellular localization,green fluorescence was visualized by confocal laser microscopy(LSM880, Zeiss, Jena, Germany).The full-length GhPdBG cDNA(without stop codon)was inserted into pGWB454.The constructed vector was agroinfiltrated into a transgenic N.tabacum line expressing CMV-MP:GFP (cucumber mosaic virus movement protein:GFP)as the PD marker[39].Yellow signal indicated that PdBG and CMV-MP:GFP were colocalized.Fiji(Fiji Is Just ImageJ,https://imagej.net/Fiji/Downloads)[40]was used to construct scatterplots of pixel signal intensities and correlations in subcellular colocalization.

2.5.Loading of carboxyfluorescein and confocal laser scanning microscopy

Following Ruan et al.[8], carboxyfluorescein (CF) was loaded into the base of fiber from underlying seed coat,and confocal imaging was used to track its entry into fibers.A 2.0%(w/v)solution of 5(6)-carboxyfluorescein diacetate(CFDA;Sigma-Aldrich,Darmstadt,Germany) dissolved in acetone was prepared and held at –20 °C.Five bolls from three plant lines were detected for each experiment.For each sample, imaging was performed on at least seven replicates per seed.

2.6.Vector construction and plant transformation

GhPdBG-D and GhBG-A were aligned with ClustalW (https://www.ebi.ac.uk/Tools/msa/clustalw2/) to identify regions suitable for virus-induced gene silencing (VIGS) and RNA interference(RNAi) fragment amplification.Total RNA from 18 DPA fibers in the cotton genetic standard line TM-1 was extracted and cDNA synthesis was performed as described above.To create overexpression (OE) and RNAi constructs, the full-length cDNA and specific,unique fragments of GhPdBG were amplified,and the PCR products were inserted into the plasmid vectors pEarlyGate 203 and pB7GWIWG2(II), respectively.Using the LR reaction [41], the p35S:GhPdBG OE and RNAi constructs were generated.Agrobacterium tumefaciens-mediated (strain LBA4404) transformation was used to introduce the OE and RNAi recombinant plasmids into cotton.

2.7.VIGS

The vector for VIGS was based on the tobacco rattle virus(TRV).The GhPdBG specific region from TM-1 for VIGS described above was inserted into the TOPO vector.The LR reaction was used to construct expression vector pTRV2-GhPdBG with p35S promoter.The steps of the VIGS assay were performed as described previously [42].Briefly, pTRV2 and pTRV1, which harbors the coat protein and the sequence used for VIGS, were introduced into Agrobacterium competent cell line GV3101 by electroporation,with empty vector pTRV-00 (pTRV1: pTRV2 = 1: 1, v/v) used as a negative control and pTRV-PDS (pTRV1: pTRV2-PDS = 1: 1, v/v) a positive control.A 5-mL culture was cultivated in LB medium containing 50 mg L-1kanamycin overnight at 28°C,and on the following day was inoculated into 50 mL LB medium containing antibiotics, 20 μmol L-1acetosyringone, and 10 mmol L-1(2-[Nmorpholino] ethanesulfonic acid (MES).The culture was grown overnight in a 28 °C shaker until the OD600 reached a value between 0.8 and 1.2.Agrobacterium cells were collected and resuspended in media(200 μmol L-1acetosyringone,10 mmol L-1MES,10 mmol L-1MgCl2), adjusted to 4.0 OD, and incubated at room temperature for 3 h.Agroinfiltration into cotton fruit, branches,and leaves was performed with a 1-mL syringe after anthesis with a mixture of the two strains in a 1:1 ratio.

2.8.Cotton quality inspection

For fiber quality measurement, fiber samples from T2generation of three OE and two RNAi GhPdBG transgenic lines,three VIGS lines, and the wild type were tested at the Institute of Cotton Research of CAAS(Anyang,Henan,China).The fibers on seeds were combed well, then gently cut off along the seed coat, spread,arranged, and counted under a stereoscopic microscope (Nikon SMZ25,JPN).Six bolls were detected for each line.Statistical analysis was based on Student’s t-test.

3.Results

3.1.Analysis of genetic variation in the GhPdBG among cotton germplasm

Plasmodesmata-mediated fiber development and evolution correspond to diverse expression patterns of the PD β-1,3-glucanase with distinct promoter variation.To identify evolutionally important specific candidate members of β-1,3-glucanase genes family that could regulate fiber PD,we screened a resequencing database of worldwide cotton germplasm resources [36] of 130 β-1,3-glucanase gene family members [35].A conserved coding region in a gene suggests that the gene is evolutionarily important, and promoter regulatory variations suggest that the gene is evolutionarily active or diversified with flexible regulation mode.The purpose of our study was to identify candidate genes whose function, was regulated by the gene expression pattern or level,but not the structure of their proteins.Based on the genetic variation analyses combined with the preferential fiber expression pattern and PD subcellular location feature in the next section, we identified the β-1,3-glucanase gene Gh_D05G3800 (designated GhPdBG due to PD localization), as a candidate gene.This gene’s coding region was completely conserved among different accessions (Fig.1A, B) but showed genetic variation in the form of single-nucleotide polymorphisms (SNPs) in its promoter region,with 11 distinct haplotypes in the germplasm panel (Fig.1C;Table S2).These 11 haplotypes showed different distributions across the ecoregion groups, with haplotype 1 accounting for the greatest proportion in all groups (Fig.1D).

Except for haplotypes 9 and 10, which lacked fiber length data[36], we compared fiber length variation among the haplotypes.Haplotype 3 containing TM-1 was associated with short fibers and haplotype 11 with long fibers (Fig.1E).These two haplotypes represent two germplasm groups with differences in fiber length from most other haplotypes (Fig.1E).Expression of the GhPdBG gene was detected in three representative haplotypes: haplotypes 1 (with the largest population), 3 (with short fibers), and 11 (with long fibers).The expression of the GhPdBG gene was highest in haplotype 11 followed by haplotype 1, and was lowest in haplotype 3(Fig.S1), the variation in gene expression was consistent with the variation in fiber length.These findings suggest that promoter variation results in variation in regulation of expression of the GhPdBG gene during fiber development.

3.2.GhPdBG was expressed predominantly in fibers and localized to plasmodesmata

qRT-PCR was used to measure the spatial expression of GhPdBG and GbPdBG.GhPdBG expression was measured in cotton genetic standard line TM-1 and GbPdBG in Hai7124 in 11 tissues: roots,stems, hypocotyls, leaves, cotyledons, petals, pistils, stamens,ovules, fibers, and sepals.The GhPdBG and GbPdBG genes were expressed predominantly in fibers (Fig.2A, B).The fiberpreferential expression pattern of PdBG suggested that the gene is involved in fiber development.

The GhPdBG ORF(without stop codon)was fused to GFP and RFP under the 35S constitutive promoter (p35S:GhPdBG-GFP, p35S:GhPdBG-RFP) for respectively subcellular localization and colocalization analyses.GhPdBG exhibited a punctate pattern on PD,and colocalized with CMV-MP:GFP (PD marker) in tobacco(Fig.2C, D).Fig.2E displays the red and green pixel signal intensities for fluorescence signal quantification scatterplots, and the green CMV-MP:GFP and red GhPdBG:RFP Pearson correlation coefficient (PCC) of the images was 0.95, confirming that GhPdBG was localized in PD.

3.3.PdBG displayed differing temporal expression patterns corresponding to intercellular permeability during fiber development in G.hirsutum and G.barbadense

The phenotypes associated with fiber quantity and fiber quality are known to differ between G.hirsutum and G.barbadense,(G.barbadense has longer fiber but lower yields than G.hirsutum).To investigate the expression pattern and intercellular activities of the PdBG gene, we characterized its temporal expression patterns by qRT-PCR and measured intercellular permeability in G.hirsutum and G.barbadense during fiber development.In G.hirsutum and G.barbadense, transcription of GhPdBG and GbPdBG

(Gbar_D05G023560) was barely detectable at the early stage of fiber development, and transcription of GhPdBG increased at 20 and 25 DPA (Fig.3A), whereas in G.barbadense, transcription of GbPdBG remained low through 0–10 DPA, increased dramatically at 15 DPA, decreased at 20 DPA, and increased again at 25 DPA(Fig.3B).Thus,the temporal expression pattern of GbPdBG differed from that of GhPdBG.

Fig.1.The polymorphic haplotype of the cotton GhPdBG promoter region.(A)The coding region of the GhPdBG gene showed unique haplotypes in all 352 accessions.(B)The proportion of haplotypes in the coding region of the GhPdBG gene in each group.(C)The proportion of different haplotypes of the GhPdBG gene promoter region across all 352 cotton accessions.(D) The proportion of different haplotypes of the GhPdBG gene promoter region in each group.(E) The fiber length trait of cotton accessions in different haplotype groups.Error bars indicate standard error of the mean (SEM).*, P < 0.05 (One-way ANOVA).

Fig.2.GhPdBG is preferentially expressed in fibers and localizes to PD.(A,B)Tissue spatial expression analysis of PdBG in G.hirsutum(TM-1)and G.barbadense(Hai7124).(A)qRT-PCR detection of GhPdBG in 11 cotton tissues.(B)qRT-PCR detection of GbPdBG in different tissues.The respective samples of ovules and fibers are separate mixtures of multiple stages: 5, 10, 15, 20, and 25 DPA.Error bars indicate standard error of the mean.(C) GhPdBG subcellular fluorescence signal localized near the PD in Nicotiana benthamiana wild-type plants.(D)GhPdBG colocalized with CMV-MP:GFP as PD marker.The leaves of CMV-MP:GFP-transgenic tobacco plants were agroinfiltrated with the C-terminal RFP-tagged GhPdBG.In merged images,yellow signal indicates colocalization of GhPdBG and CMV-MP:GFP.Confocal laser scanning microscopy(CLSM)was used to obtain images.Scale bars,20 μm in(C)and(D).(E)CMV-MP:GFP and GhPdBG:RFP are fluorescence signal quantification scatterplot of red and green pixel intensities.The Pearson correlation coefficient (PCC) of the red GhPdBG:RFP and green CMV-MP:GFP in images was 0.95.DPA, days post-anthesis.

To directly observe the gating state of PD, PD permeability fluorescence was measured in G.hirsutum and G.barbadense.Loading carboxyfluorescein diacetate (CFDA) at the base of fiber from underlying seed coat is an effective means of probing PD status and the connectivity of fiber cells.At 10 DPA,the PD gates in G.hirsutum and G.barbadense were closed (Fig.3C, G).A marked difference between G.hirsutum and G.barbadense was observed at 15 and 20 DPA; at 15 DPA, PD were closed in G.hirsutum but open in G.barbadense, and at 20 DPA, PD were open in G.hirsutum but closed in G.barbadense (Fig.3C–J).The results of the PdBG transcription level analyses (Fig.3A–B) and PD permeability fluorescence experiments (Fig.3C–J) show that in both species, low PdBG transcript levels were accompanied by PD gate closure,whereas high PdBG transcript levels were accompanied by PD gate opening.Comparing the full gene sequences of GhPdBG and GbPdBG in G.hirsutum and G.barbadense showed that the coding sequences were highly conserved, while the promoter regions differed, especially the 1.2 kb promoter region near the initiator codon(Fig.S2).This finding suggests that PdBG expression is differently regulated in G.hirsutum and G.barbadense, possibly owing to genetic variation in the promoter.Thus, expression of the PdBG gene is associated with the opening and closing of PD.

3.4.GhPdBG functions in regulating fiber and trichome development

Fig.3.Differing expression patterns and permeability of PD during fiber development in G.hirsutum and G.barbadense.(A) and (B) qRT-PCR detection of PdBG temporal expression in fibers of G.hirsutum and G.barbadense at four developmental stages.Error bars indicate standard error of the mean (SEM).(C–J) Fiber PD permeability during development in G.hirsutum and G.barbadense.Fluorescence signal in fiber,implying open PD,was observed in G.hirsutum at 20 DPA(E)and 25 DPA(F),and in G.barbadense at 15 DPA (H) and 25 DPA (J); at the other timepoints, no fluorescence signal was observed in fiber, implying closed PD.Five bolls from three plant lines were analyzed for each experiment.Scale bars, 400 μm in (E–L).DPA, days post-anthesis.

In addition to observing native expression patterns and regulation modes found above,we wished to further investigate the function of GhPdBG by manipulating its expression.GhPdBG gene expression was reduced via VIGS.Visual inspection showed that the 20-DPA cotton bolls of VIGS interference plants were smaller than those of empty-vector transgenic plants pTRV-00.The mature fiber of VIGS was shorter than that of the pTRV-00 (Fig.4A–C),while the size and number of seeds in bolls were unchanged, and as mentioned above, PdBG was negligibly expressed in ovules.qRT-PCR analysis revealed that expression of GhPdBG in VIGS plants remained nearly unchanged at 15 d but decreased at 20 d compared with that in wild-type(WT)mock plants(Fig.4D).Thus,GhPdBG expression was associated with fiber and boll development.

To further investigate the function of GhPdBG, overexpression(OE)and RNA interference(RNAi)vectors of GhPdBG gene from cotton genetic standard line TM-1 were transformed into cotton using Agrobacterium-mediated transformation.Gossypium hirsutum transgenics were identified based on Bar gene PCR screening(data not shown),and qRT-PCR was used to evaluate GhPdBG expression in OE and RNAi transgenic lines.Besides showing lower expression of GhPdBG (Fig.5A), RNAi lines had shorter fiber and stem trichomes and smaller bolls than WT (Fig.5B–F).But the size and number of seeds in bolls were unchanged, in agreement with the phenotypes produced by VIGS (Fig.4A–B).In contrast, fiber and stem trichomes in the GhPdBG OE line were longer than those of WT, and fiber density was also greatly increased.The number of cotton fibers and boll size of OE transgenic lines were higher than those of WT.There is no phenotypic difference showed no visible in other main organs,as shown with the AtPdBG4 gene in Arabidopsis [33].

Thus,a decrease in GhPdBG gene expression led to reductions in length of seed fibers and stem trichomes.Overexpression of GhPdBG increased both quantity and length of fibers and trichomes.

3.5.GhPdBG functions by modulating plasmodesmatal permeability

To further investigate the cellular physiological basis and verify the phenotypic alteration in fiber development observed in GhPdBG transgenic plants, fluorescence-based fiber PD permeability experiments were conducted in transgenic plants using CFDA.In WT as control, no trafficking of fluorescent dye was observed in fibers at 10 or 15 DPA; however, fluorescent dye was transported to fibers at 20 DPA(Fig.3C–E).In OE transgenic lines,entry of fluorescent dye into fibers was observed at 10, 15, and 20 DPA,indicating that PD in OE plants were opened in the 10 to 20 DPA period(Fig.5G–I).In contrast,in RNAi transgenic lines,no fluorescent dye was detected in fibers at 10, 15, or 20 DPA (Fig.5J–L).These results support the hypothesis that ectopic expression of GhPdBG regulates fiber PD permeability.

4.Discussion

4.1.PD opening pattern may determine interspecific phenotypic variation in cotton fibers

Fig.4.Phenotypes of GhPdBG VIGS interference lines.(A) Cotton bolls of pTRV-00 and GhPdBG VIGS interference plants at 20 DPA.Compared to pTRV-00 plants, VIGS interference plants’cotton bolls were smaller;V,VIGS.(B)Mature fibers of pTRV-00 and VIGS plants.The mature fibers of VIGS plants were shorter than those of the pTRV-00 plants.(C)VIGS plants have shorter fiber lengths than pTRV-00 plants.(D)qRT-PCR expression analysis of GhPdBG between WT and VIGS plants.***,P<0.001(Student’s t-test).Error bars indicate standard error of the mean.DPA, days post-anthesis.

Based on genetic variation among cotton accessions,we identified a member of the β-1,3-glucanase family, PdBG, in which the coding regions were completely conserved but the promoter regions showed marked differences.Subcellular localization confirmed that GhPdBG is a plasmodesmata-specific gene.During fiber initiation and elongation,the PdBG gene displayed distinct expression patterns.The PD of WT G.hirsutum were closed at 10 and 15 DPA and open at 20 and 25 DPA, a finding consistent with that of Ruan et al.[8].The PD open and close states of G.barbadense at 15 and 20 DPA were consistent with those previously reported[8].The finding that at the early stage of fiber elongation (before 10 DPA) in both G.hirsutum and G.barbadense, transcription was negligible indicates that PdBG was not expressed in this period and that the PD were closed.This finding differs from previous finding [8] that the PD were open before 10 DPA, possibly owing to species specificity and/or time-dependent expression mode during the day.

In view of the sharp differences in expression of PdBG in G.hirsutum and G.barbadense, we suggest the following explanation.The promoter regions of GhPdBG and GbPdBG varied between the two species.The difference in expression patterns between the two species may have been due to the action of different expression regulatory factors (transcription factors, methylation factors,microRNA, etc.) in the two genetic backgrounds.Differing regulation of PdBG expression in G.hirsutum and G.barbadense could be attributed to genetic variation in the promoter and related regulatory factors.

The different PdBG expression and PD opening patterns in the two species may contribute to the differences in the fiber traits of G.barbadense and G.hirsutum.

4.2.Excessive opening of plasmodesmata promotes both the initiation and elongation of fibers

Although β-1,3-glucanase proteins are known to function in processing callose [25], we wished to investigate the molecular basis of its regulation and its functional consequences.The PdBG expression was found to mediate the opening and closure of PD during fiber development in WT G.hirsutum and G.barbadense.The finding that GhPdBG OE and RNAi dramatically altered the length and/or quantity of fibers and stem trichomes suggested that seed fiber and stem trichome cell development are dependent on GhPdBG gene expression.The finding in GhPdBG OE transgenic plants that excessive opening of PD increased the number and length of seed fibers and stem trichomes suggested that PD opening promoted the initiation and elongation process of both seed fibers and stem trichomes.

Excessive opening of PD leads to unexpected of promoting cotton fiber and trichome development.The phenotypes may have been caused by differences in genetic background,gene expression,intercellular communication, and physiological substance exchange with an unknown novel developmental mode.Fiber elongation is a type of plant cell growth and requires a series of substances such as nutrients and signals.Many substances are transported through PD[43].The opening of PD could be conducive to material transport and could promote cell elongation and growth.The results suggest that PD’s excessive opening mode, as another regulatory determinant with a positive promoting effect at the early fiber development stage, is conducive to cotton trichome and fiber cell development.

AtPdBG4 is a homolog gene of GhPdBG[33].Phenotypic was analyzed on the overexpression and RNA interference lines of GhPdBG and AtPdBG4 in cotton and Arabidopsis, respectively.The functional features of PdBG differed between species.We propose that the various trichomes possess organ- and species-specific regulation modes and pathways in the biological microenvironment during trichome development.This model suggests that the PdBG gene has evolved to diversely regulate trichome cell initiation and elongation during this complex and concerted developmental process in higher plants.

Fig.5.Phenotypes and fiber PD permeability in ovule epidermis of GhPdBG OE and RNAi lines.(A)Expression measurement of GhPdBG in representative fiber samples from OE-21,RNAi-27,and CK plants.(B)Bolls of OE-21,RNAi-27,and WT plants at 20 DPA.(C)Mature fibers of OE-21,RNAi-27,and WT plants.(D)Comparison of the phenotypes of stem trichomes of OE-21 and RNAi-27 lines with those of WT cotton.The images in the second row show enlarged views of portions in red frames.(E)OE transgenic line contained more lint fibers than WT,but RNAi transgenic line contained fewer lint fibers than WT.The number of lint fibers on seed epidermis was greater in OE transgenic line than in WT line,and there was no change between RNAi line and WT line.Six bolls from four plants were examined for each line.(F)The fiber of OE line was longer than that of WT, and that of RNAi line was shorter.Six bolls from four plants were examined for each line.Three GhPdBG OE and two GhPdBG RNAi transgenic lines, with six biological repeats were detected.*,P<0.05;**,P<0.01(Student’s t-test).Error bars indicate standard error of the mean.(G–L)Carboxyfluorescein(CF)dye is transported via basal plasmodesmata(PD)from the ovule epidermis to the fibers in transgenic plants.Fluorescence of fiber cells on developing ovules of transgenic(OE,RNAi)plant at 10,15,and 20 DPA.(G–I)In OE transgenic plants,CF accumulated in fibers at all stages,that is,at 10,15,and 20 DPA,indicating that PD was open.(J–L)In RNAi transgenic plants,no CF was detected in fibers at 10,15,and 20 DPA,indicating that PD were closed.Scale bars,400 μm in(G–L).As control in WT plants,PD permeability results were shown in Fig.3C–E; CF was transported into fibers at 20 DPA, indicating open PD,while signals were not detected at 10 and 15 DPA, indicating closed PD.Five bolls from three plant lines were examined for each experiment.DPA, days post-anthesis.

4.3.Reducing the expression of GhPdBG inhibits fiber elongation but does not affect cotton fiber number

In transgenic cotton plants,both the number and length of cotton fibers increased when the expression of the GhPdBG gene increased.Although the length of cotton fibers decreased when the expression of GhPdBG was suppressed, the fiber number on the seed epidermis did not change.Excessive closing of PD altered fiber length but not fiber number in RNAi plants,indicating that its effect was a result of inhibition of cotton fiber elongation but not initiation.We suggest two explanations for the differing effects on elongation and initiation.The first is that although the GhPdBG gene functions in fiber elongation, some other backup pathways can also affect fiber initiation and may partially compensate for the downregulation of GhPdBG.The second is that the expression of GhPdBG was downregulated by RNAi rather than completely silenced; thus, the PD did not close to a sufficient degree below the threshold to affect the developmental phenotype.

These findings about the GhPdBG functional phenotype appear to be contrary to the conventional view.We have no information about the physiological mechanism involved in fiber cell development under constitutively open or closed PD mediated by ectopic regulation of GhPdBG (via gene overexpression and knockdown).However,we speculate that the PdBG-mediated PD mode is a novel mechanism for the regulation of intercellular channels to control trichome development.

4.4.Further study and utilization of the PdBG gene for investigating physiological mechanisms and germplasm development

This study identified and characterized a functional switch controlling intercellular PD for fiber development in cotton.The knowledge provides an insight into intercellular transport on trichome development, which is of fundamental importance for unraveling the nature of single-cell development in higher plants.

Although our findings provide clues to the physiological process regulating intercellular information exchange and molecular trafficking through PD, future studies on this subject related to PdBG would elucidate the regulatory mechanism of trichomes development process as a model for cell differentiation.

We propose that PdBG represents a functional factor affecting cotton fiber development and final traits.We envisage that future practical application using gene editing technology could modify the gene promoter to create novel beneficial germplasm, improving fiber quantity and quality by increasing the number and length of fibers for textile production, as well as trichome-based resistance to both biotic and abiotic stresses.

Accession numbers

The sequence accession numbers for this article can be found at CottonGen(https://www.cottongen.org/,June 15,2023)as follows:GhPdBG, Gh_D05G3800; GbPdBG, Gbar_D05G023560.

CRediT authorship contribution statement

Yijie Fan:Data curation,Investigation,Formal analysis,Writing– original draft, Writing – review & editing.Shuangshuang Lin:Investigation, Formal analysis, Writing – original draft.Yanhui Lyu:Investigation,Formal analysis,Methodology.Haihong Shang:Funding acquisition, Project administration, Methodology.Youlu Yuan:Funding acquisition, Project administration.Zhengmin Tang:Resources, Software, Formal analysis.Chengzhi Jiao:Methodology, Software.Aiyun Chen:Investigation.Piyi Xing:Methodology, Software.Li Zhang:Methodology, Writing – review& editing.Yuxiao Sun:Methodology.Haixia Guo:Methodology,Visualization.Tongtong Li:Data curation,Methodology.Zhonghai Ren:Supervision, Methodology.Fanchang Zeng:Funding acquisition, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was supported by the State Key Laboratory of Cotton Biology Open Fund (CB2021A04), the Agricultural Seed Project of Shandong Province (2020LZGC002), and the Science Foundation of Shandong Province (ZR2020MC107).Acknowledgments to Dr.Yinhua Jia and Dr.Xiongming Du in the germplasm center of Institute of Cotton Research of the Chinese Academy of Agricultural Sciences (CAAS) for providing the seeds of germplasm material.

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2023.06.010.