A polygalacturonase gene OsPG1 modulates water homeostasis in rice

Qinwen Zou, Rnrn Tu, Jijun Wu, Tingting Hung, Zhiho Sun, Zheyn Run,Hongyu Co, Shihui Yng, Xihong Shen,*, Gunghu He,*, Hong Wng,*

a Rice Research Institute, Key Laboratory of Application and Safety Control of Genetically Modified Crops, Academy of Agricultural Sciences, Southwest University, Chongqing 400715, China

b State Key Laboratory of Rice Biology and Breeding, Key Laboratory for Zhejiang Super Rice Research, China National Center for Rice Improvement, China National Rice Research Institute, Hangzhou 311401, Zhejiang, China

Keywords:Photosynthesis Stomata Transpiration Leaf rolling

ABSTRACT A dynamic plant architecture is the basis of plant adaptation to changing environments.Although many genes regulating leaf rolling have been identified, genes directly associated with water homeostasis are largely unknown.Here, we isolated a rice mutant, dynamic leaf rolling 1 (dlr1), characterized by ‘leaf unfolding in the morning-leaf rolling at noon-leaf unfolding in the evening’ during a sunny day.Water content was decreased in rolled leaves and water sprayed on leaves caused reopening, indicating that in vivo water deficiency induced the leaf rolling.Map-based cloning and expression tests demonstrated that an A1400G single base mutation in Oryza sativa Polygalacturonase 1 (OsPG1)/PHOTO-SENSITIVE LEAF ROLLING 1 (PSL1) was responsible for the dynamic leaf rolling phenotype in the dlr1 mutant.OsPG1 encodes a polygalacturonase, one of the main enzymes that degrade demethylesterified homogalacturonans in plant cell walls.OsPG1 was constitutively expressed in various tissues and was enriched in stomata.Mutants of the OsPG1 gene exhibited defects in stomatal closure and decreased stomatal density,leading to reduced transpiration and excessive water loss under specific conditions,but had normal root development.Further analysis revealed that mutation of OsPG1 led to reduced pectinase activity in the leaves and increased demethylesterified homogalacturonans in guard cells.Our findings reveal a mechanism by which OsPG1 modulates water homeostasis to control dynamic leaf rolling, providing insights for plants to adapt to environmental variation.

1.Introduction

Leaf morphology, an important component of plant architecture, directly affects photosynthetic efficiency, accumulation of dry matter, and thus grain yield [1].Leaf rolling, coordinately regulated by genetic and environmental factors[2],is a dynamic trait,because plants require constant adjustment in responding to a changing environment [3].Temporary leaf rolling is a common reversible water stress response [4,5].Elucidation of the mechanisms underlying dynamic leaf rolling will contribute to an understanding of how plants adapt to environmental change.

Leaf rolling is genetically influenced by genes regulating specific cells or structures, including bulliform cells, sclerenchymatous cells, and the leaf cuticle [1].Bulliform cells, also known as motor cells, are a class of large, highly vacuolated cells located in adaxial leaf epidermis in monocotyledons [6].Change in bulliform cell number or size can affect leaf rolling.Many genes affecting size or number of bulliform cells in rice have been identified.For example,overexpression of Abaxially Curled Leaf 1 (ACL1) or its homologous gene ACL2 in rice caused abaxial leaf rolling associated with increased number and size of bulliform cells [7].Overexpression of OsZHD1 encoding a zinc finger homeobox transcription factor increased the number of bulliform cells in the paraxial cell layer of rice leaves, leading to abaxial leaf rolling [8].A transcriptional repression complex UPWARD ROLLED LEAF 1(URL1)/RICE OUTERMOST CELL-SPECIFIC 8 (ROC8)-ROC5-TPL2 suppresses ACL1 expression to control leaf rolling through negatively regulating the number and size of bulliform cells [9,10].OsSNF7.2, a component of ESCRT-III,regulated the number and size of bulliform cells to control adaxial leaf rolling by trafficking and endosomal degradation of auxin biosynthetic enzyme OsYUC8 [11].Mutation of SEMI-ROLLED LEAF 10(SRL10)increased bulliform cell number and size and caused adaxial leaf rolling [12].

Water content in vivo can alter bulliform cells size and thus control leaf rolling.When plant leaves lack water, bulliform cells shrink, leading to leaf rolling; When plant leaves have sufficient water, the bulliform cells absorb water and expand, unfolding the leaves [5].This indicates that leaf rolling can change with in vivo water content.Three genes in rice, a 2OG-Fe(II) oxygenase gene named RL14 (ROLLING-LEAF14), a CURLED LEAF AND DWARF1 gene named CLD1 (or SEMI-ROLLED LEAF 1 (SRL1)), and CONSTITUTIVELY WILTED1 (OsCOW1), have been identified to regulate in vivo water content, and loss of function of these genes causes permanent leaf rolling rather than reversible or dynamic leaf rolling [13–16].Overexpression of ROOT UV-B SENSITIVE1 (OsRUS1)and mutation of PHOTO-SENSITIVE LEAF ROLLING 1(PSL1)encoding a polygalacturonase led to dynamic leaf rolling caused by light[17,18], but whether these mutants affect in vivo water content remains to be elucidated.

Polygalacturonase(PG)is one of the main degrading enzymes of pectins that are abundant, branched polysaccharides present in plant cell walls.PG is a GH28 hydrolase family that modifies plant cell walls and affects cell wall integrity by hydrolyzing demethylesterified homogalacturonan (HG), a pectin component[19–21].PG functions in multiple processes during plant growth and development.PG contributes to fruit softening, and antisense suppression of PG genes, such as FaPG1 in strawberry [22],MdPG1 in apple [23], and PpPG1/2 in peach [24], produce firmer textured fruits due to the reduced pectin polymer depolymerization and solubilization,as well as cell-to-cell separation[21].PG genes in Arabidopsis thaliana are involved in cell division and expansion.For example, in Arabidopsis, Dehiscencezone Polygalacturonase 1(ADPG1)and ADPG2 regulate cell separation and expansion to control floral organ abscission and silique dehiscence [25] and POLYGALACTURONASE INVOLVED IN EXPANSION1 (PGX1) and PGX2 regulate cell expansion to control hypocotyl elongation [26,27].In rice,PG affects cell wall integrity and cell adhesion.Overexpression of OsBURP16 encoding a PG1β subunit caused reduced pectin content and cell adhesion, as well as transpiration rate, leading to decreased tolerance to cold, salinity, and drought stresses [28].Mutation of OsPG1/PSL1 altered the thicknesses of cell walls and cell wall composition, and caused leaf tip necrosis, enhanced bacterial blight resistance, and photo-sensitive leaf rolling [18,29].

In this study,we identified a dynamic leaf rolling 1(dlr1)mutant showing reversible leaf rolling with maximum leaf rolling at noon on sunny days.Water sprayed on leaves rescued the rolled leaf phenotype, suggesting that in vivo water deficiency caused leaf rolling in dlr1 mutant leaves.An A1400G single base mutation in OsPG1/PSL1 that encodes a PG caused the dynamic leaf rolling phenotype,which led to decreased pectinase activity in the leaves and increased demethylesterified HGs in guard cells,affecting stomatal density and activity and thus in vivo water homeostasis.Our findings provide insights into the mechanisms underlying dynamic leaf rolling in rice.

2.Materials and methods

2.1.Plant material and growing conditions

The dlr1 mutant was selected following ethyl methanesulfonate(EMS)mutagenesis of japonica rice variety Changgeng 3(CG3,wild type, WT).An F2population was generated by crossing the dlr1 mutant with indica cultivar Nanjing 11.All plant materials were grown in the paddy fields at the Rice Research Institute of Southwest University in Chongqing.

2.2.Agronomic traits

Ten plants from WT and the dlr1 mutant,respectively,were randomly harvested at maturity and agronomic traits plant height,tiller number, panicle length, number of filled grains per panicle,seed setting rate, and 1000-grain weight were measured.Twotail Student’s t-tests were used in statistical analyses.

2.3.Fine mapping of the OsPG1 locus

Bulked segregant analysis (BSA) was used for linkage analysis.Three hundred and ninety-five F2individuals with the recessive phenotype were used for fine mapping of the OsPG1 locus.New polymorphic insertion/deletion (InDel) marker development and PCR amplification were performed as previously described [30].The primers for gene mapping and sequencing are listed in Table S1.

2.4.Histology and microscopy

Leaf and root samples from the WT and the dlr1 mutant were collected and fixed in FAA solution (70% (v/v) ethanol, 5% (v/v)formaldehyde, and 5% (v/v) acetic acid) for 48 h.The materials were dehydrated in an ethanol series at increasing concentrations(50%(v/v),70%(v/v),85%(v/v),95%(v/v),and 100%(v/v)),and then infiltrated with xylene and embedded in paraffin (Sigma, St Louis,MO, USA).The samples were subsequently microtomed into 10-μm thickness sections, and mounted on poly-L-Lys-coated glass slides,de-paraffinized with xylene,and finally rehydrated through an ethanol series of reducing concentration.The sections were stained with Fast Green (Amresco, Framingham, MA, USA) and counterstained with Safranin (Amresco), dehydrated through an ethanol series,cleared with xylene,and finally mounted in neutral balsam and covered with a coverslip.Light microscopy of the sections was performed using an Eclipse E600 microscope (Nikon,Tokyo).Cell size was measured with ImageJ (National Institute of Health, https://imagej.net/ij/index.html).

2.5.Propidium iodide and GUS staining

Roots from plants at the grain filling stage were stained with propidium iodide solution (Sigma-Aldrich, Beijing) as previously described [31].Root, stem, leaf, sheath, and panicle tissue from transformed plants expressing pOsPG1::GUS were sampled and incubated in darkness for at least 12 h at 37°C using a GUS staining solution(50 mmol L-1PBS buffer;10 mmol L-1EDTA,pH 8.0;0.1%(v/v) Triton X-100; 1 mg mL-1X-gluc; 1 mmol L-1potassium ferricyanide; 1 mmol L-1potassium ferrocyanide).Chlorophyll was removed by boiling in 95% ethanol.

2.6.Scanning electron microscopy (SEM)

The shape and number of stomata in the same leaf region of WT,dlr1 mutant, knockout mutant 1/2 (ko-1/2), and the complementary (COM) line were observed by a scanning electron microscope(Hitachi SU3500, Japan) under strong vacuum.The stomatal apertures of three plants per genotype were determined,and 40 stomata were measured for each plant.The numbers of stomata in an area of 24.26 mm2were counted.

2.7.Vector construction and plant transformation

A 5984-bp fragment containing the OsPG1 genomic sequence was cloned into binary vector pCAMBIA1300 to generate the complementation vector pCAMBIA1300-OsPG1.For gene overexpression, the 1512-bp CDS of the OsPG1 allele was cloned into the binary vector pCAMBIA2300 to generate overexpression vector pCAMBIA2300- OsPG1 under the control of the Actin1 promoter.To generate gene-knockout constructs,a target sequence in OsPG1 was designed using the online CRISPR-P tool(https://cbi.hzau.edu.cn/cgi-bin/CRISPR) and then inserted into the pcas9-sgRNA-AarI backbone under the control of the OsU3 promoter.For pOsPG1::GUS vector construction, a 2448-bp genomic fragment upstream of the OsPG1 start codon was amplified and cloned into the pCAMBIA1305 vector.The above vectors were transformed into the corresponding background by Agrobacterium tumefaciens-mediated transformation to generate complementation, overexpression,gene knockout, and GUS reporter transgenic lines.The primer sequences used for vector construction are listed in Table S2.

2.8.RNA extraction and quantitative real-time PCR (qRT-PCR)

Total RNA was extracted from rice samples using RNAprep Pure Plant RNA Purification kit (Tiangen, Beijing).cDNA synthesis and qRT-PCR were performed as previously described[32].The rice Actin gene (LOC_Os03g50885) was used as an internal control, and the 2-ΔΔCTmethod was used to calculate relative gene expression.Each experiment was performed with three biological replicates.The gene-specific primers used for qRT-PCR are listed in Table S3.

2.9.Physiological measurements

Photosynthesis, stomatal conductance, and transpiration were measured at tillering using a steady-state porometer (Li6400,LICOR Biosciences, Lincoln, NE, USA).Three individuals were selected for each line.To determine the dynamics of leaf rolling on a sunny day, the leaf rolling index (LRI) was measured every hour from 7:00 to 19:00.The widths of the flag leaves of plants at tillering were measured under the natural (Ln, changing during a sunny day)or(Lw,the real leaf width)state.LRI was calculated as‘LRI = (Lw - Ln)/Lw’.The water content of leaves was determined as ‘water content (%) = (M1 - M2)/M1 × 100%’ (M1 indicates the fresh leaf weight, and M2 indicates the leaf dry weight after 15 h of drying at 60 °C).Detached leaves of WT and dlr1 mutants were placed in an illuminated growth chamber (RDN-1000B, YANGHUI,Ningbo, Zhejiang, China),and measurements of rates of water loss were performed as previously described [33].

2.10.Measurement of pectinase activity

Leaves of 50-day-old WT, dlr1 mutant, ko-1/2, and COM plants were ground to fine powder in liquid nitrogen,and ～0.1 g powder was used for each line.Pectinase activity was measured using a pectinase kit (A140-1-1, Nanjing Jiancheng Bioengineering Institute, Nanjing, Jiangsu, China).Pectinase hydrolyzes pectin to generate galacturonic acid, which has a reducible aldehyde group and reacts with 3,5-dinitrosalicylic acid (DNS) reagent to generate a reddish-brown substance with a characteristic absorption peak at 540 nm;pectinase activity was calculated based on the absorbance value at 540 nm.

2.11.Immunolabeling

To detect demethylesterifed homogalacturonan(HG),longitudinal sections of rice leaves were incubated overnight with the primary antibody JIM5 (AS184194, Agrisera) at 1:10 dilution in a phosphate buffer (6 mmol L-1Na2HPO4, 3.5 mmol L-1KH2PO4,2.7 mmol L-1KCl, pH 7.4) at 4 °C.Sections were then washed and incubated with a Goat anti-Mouse IgG(H+L)Fluor 594(Affinity) at 1:100 dilution for 1 h at 37 °C, washed again, and observed using a confocal microscope (LSM800, Zeiss).A wavelength of 561 nm was used for signal detection.

3.Results

3.1.The dlr1 mutant exhibits the dynamic leaf rolling phenotype

The leaf rolling behavior of the dlr1 mutant was compared to the WT.In the morning of a sunny day (temperature, relative humidity,and light intensity are shown in Fig.S1),the dlr1 mutant displayed flat leaf blades similar to the WT(Fig.1A–C);at noon,the dlr1 mutant exhibited adaxially rolled leaves compared with flat leaves on WT plants (Fig.1D–F); in the evening, the rolled leaves of dlr1 plants returned to the flat state similar to the WT(Fig.1G–I).Leaf rolling indices (LRI) were determined at intervals from 7:00 to 19:00.The LRI of dlr1 mutant leaves was the lowest at ～5.9% at 7:00, reached a maximum of ～83.1% at 14:00, and decreased to ～6.5% at 19:00, whereas the LRI of WT leaves fluctuated within a narrow range from ～1.0% to ～9.6% (Fig.1J).Older leaves of dlr1 mutant plants at the tillering stage exhibited distinct leaf tip necrosis (Fig.1K, L).Plant height, panicle length, panicle number per plant, total grain number, filled grain number, seedsetting rate, 1000-grain weight,and yield per plant of dlr1 mutant plants were significantly lower than WT plants (Fig.S2).These observations showed that dlr1 mutant plants had reduced plant characteristics compared to the wild type.

3.2.Rolled leaves from the dlr1 mutant show in vivo water deficiency

Since the dlr1 mutant displayed the leaf rolling at time points with relatively low humidity (Fig.S1) and leaf tip necrosis resembling phenotypic characteristics of rice plants under drought conditions, we speculated that the abnormal leaf rolling at noon was caused by in vivo water deficiency.Measurements of water content in WT and dlr1 mutant leaves in the morning and at noon showed that the water contents of WT and dlr1 mutant leaves in the morning were not different,but the water content in dlr1 mutant leaves was significantly decreased at noon (Fig.2A).When water was sprayed to the intact leaves of mutant and WT plants at noon,the leaves of the mutant unrolled (Fig.2B, C, F).As in vivo water content can change the size of bulliform cells that control dynamic leaf rolling [5], we then investigated the bulliform cells.The bulliform cells of dlr1 mutant leaves showed a reduced size compared with those of WT leaves before the water spray (Fig.2D, G), but they returned to a similar size to the WT after the water spray(Fig.2E,G).As the number of bulliform cells also affects leaf rolling,we measured the numbers of bulliform cells in WT and dlr1 mutant leaves but found no marked difference between the two lines(Fig.2H).These results indicate that the rolled leaves of the dlr1 mutant do indeed lack water in vivo.

3.3.Map-based cloning of the OsPG1 gene

To test whether the dynamic leaf rolling in the dlr1 mutant was controlled by a single gene,800 F2plants from a cross between the dlr1 mutant and indica cultivar Nanjing 11 were phenotyped.The ratio of 621 normal: 179 leaf-rolled individuals fitted expectation for segregation at a single locus (621:179, χ23:1= 2.94;P1df> 0.05), suggesting that dynamic leaf rolling in the dlr1 mutant was conferred by a single recessive allele.

Linkage analysis based on a bulked DNA sample from 22 F2individuals located the OsPG1 locus in a ～2.8-Mb genomic region on the short arm of chromosome 1 flanked by simple sequence repeat(SSR)markers RM600 and RM493 (Fig.3A).Further genotyping of 373 F2individuals with the recessive phenotype with 11 new polymorphic insertion/deletion (InDel) markers placed the locus in a 46.9-kb region flanked by markers R-22 and R-20 (Fig.3B).There were 5 open reading frames (ORFs) in that interval based on the Nipponbare genomic database RAP-DB (https://rapdb.dna.affffrc.go.jp/)(Fig.3C).Sequencing of the 5 ORFs identified an A1400G substitution in ORF2 in the dlr1 mutant(Fig.S3A),causing a Q400R change in the encoded protein(Fig.3C).ORF2(LOC_Os01g19170)was previously identified as OsPG1/PHOTO-SENSITIVE LEAF ROLLING 1 (PSL1)encoding a polygalacturonase (PG) [18,29].

Fig.2.Water content measurements and water spray treatments.(A) Leaf water contents of WT and dlr1 mutant leaves in the morning and at noon on a sunny day.Each measurement was performed with three biological replicates(n=3).(B,C) Leaf morphologies of WT(B)and dlr1 mutant(C)plants before and after water spray.Scale bars,5 cm.(D,E)Cross-sections of WT(D)and dlr1 mutant(E)leaves before and after water spray.Scale bars,50 μm.(F)Leaf rolling indices of WT and dlr1 mutant leaves before and after water spray.Each measurement was performed with 10 biological replicates(n=10).(G)Bulliform cell size in WT and dlr1 mutant leaves.(H)Numbers of bulliform cells flanked by two vascular bundles in WT and dlr1 mutant leaves.Each measurement was performed with 10 biological replicates (n = 10).Different letters indicate a statistical difference at P ≤0.01 by t-tests.Data are means ± SD.

Fig.3.Fine mapping of the OsPG1 locus and transgenic analysis of the OsPG1 gene.(A)Preliminary mapping placed the OsPG1 locus in to a 2.8-Mb region in the short arm of chromosome 1.(B)The OsPG1 locus was fine mapped to a 46.9-kb region.(C)Candidate gene and sequencing analysis.The blue arrows indicate the mutant site and change from glutamine to arginine.(D–I)Leaf phenotypes of WT(D),dlr1 mutant(E),COM(F),ko-1(G),ko-2(H),and OE-1(I)plants in the morning,noon and evening of a sunny day.Scale bars, 5 cm.

Genetic complementation tests showed recovery of WT phenotype verifying that the mutation in ORF2 caused the dynamic leaf rolling in the dlr1 mutant (Figs.3D, F, S3B).In addition, ORF2 knockout lines (ko-1/2) produced by CRISPR/Cas9 displayed the mutant phenotype (Figs.3E, G, H, S3C) whereas the phenotypes of overexpression lines were normal (OE-1) (Figs.3I, S3D).These results demonstrated that the A1400G mutation in ORF2 of OsPG1 caused the dynamic leaf rolling phenotype in the dlr1 mutant.

3.4.OsPG1 is constitutively expressed

Relative transcript levels quantified by qRT-PCR in various tissues at heading showed that OsPG1 was mainly expressed in the roots, leaves, leaf sheaths and panicles, but there was relatively low expression in stems (Fig.4A).To further investigate OsPG1 expression,we generated transgenic lines expressing the GUS reporter driven by the native OsPG1 promoter.Consistent with the results of qRT-PCR, GUS signals were observed in all examined tissues,including the root, stem, leaf sheath, and panicle (Fig.4B–G).

Considering that the roots are mainly responsible for water absorption and that the leaves are mainly responsible for transpiration, we further examined OsPG1 expression at the cellular level in roots and leaves of transgenic plants expressing pOsPG1::GUS in paraffin sections.GUS signals were observed in all root cell types,including the epidermis, exodermis, cortex, endodermis, phloem,and xylem (Fig.4H–J); GUS signals in the leaves were mainly detected in mesophyll cells (Fig.4K,) and were enriched in the stomata (Fig.4L).

3.5.The OsPG1 mutant lines showed normal root development

Fig.4.Expression analysis of the OsPG1 allele.(A) Transcript levels of OsPG1 in various tissues of WT plants at heading, including root, stem, leaf sheath, and panicle.GUS staining: (B–G) root (B, C), stem (D), leaf (E), leaf sheath (F), and panicle (G) of transgenic plants expressing pOsPG1::GUS.Scale bars (B, C), 1 mm; (D–G), 1 cm.(H–J)Longitudinal section(H)of a root(scale bar,1 mm)and transverse section(I,J)of the root(scale bars,50 μm).(K)Leaf section(scale bars,50 μm).(L)is an enlarged view of E;arrows indicate stomata.Scale bars, 50 μm.Data are means ± SD.

As the rolled leaves from the dlr1 mutant showed in vivo water deficiency, we attempted to determine if the OsPG1 gene was involved in the root development or water uptake.We found no differences in root length in WT, dlr1 mutant, ko-1/2 lines, and COM plants at the seedling and tillering stages (Fig.5A–C), nor were there differences in number and length of lateral roots at the seedling stage (Fig.5D, E).The xylem is composed of interconnected tracheary elements(TEs)that create long-distance capillary channels for transporting water and minerals from the roots to the aerial parts of the plant[34].Analysis xylem structures in the roots of WT, dlr1 mutant, ko-1/2, and COM plants at the grain filling stage also revealed no differences (Fig.5F–O).These observations indicated that mutation of OsPG1 had no effect on root length,lateral roots, or root xylem formation.

3.6.The OsPG1 mutant lines exhibit decreased stomatal density and defective stomatal closure

Given that stomata are the exit points for plant transpiration and that OsPG1 was expressed in the stomata, we then focused on the stomata and stomatal activity.Stomatal apertures evaluated as completely open,partially open,and completely closed(Fig.6A)indicated that in the morning the WT, dlr1 mutant, and COM leaves displayed similar stomatal apertures with more than 80% in the partially open state and the remainder completely closed (Fig.6C).At noon, WT and COM leaves showed similar stomatal apertures of 21.7%and ～25.0%, respectively, partially open and the remainder completely closed,whereas more than 95%of stomata in the dlr1 mutant and ko-1/2 leaves were partially open, and less than 5% were completely closed (Fig.6D).In the evening, the partially open stomata of WT and COM leaves were ～34.1% for the WT stomata ～37.5% for COM with the remainder completely closed, whereas all stomata in the dlr1 mutant and ko-1/2 leaves were partially open(Fig.6E).To test whether the defect in stomatal closure was associated with in vivo water deficiency, we investigated the water loss rates of detached leaves of WT and dlr1 mutant plants under relatively mild conditions(e.g., the morning of a sunny day) and relatively extreme conditions(e.g.,at noon of a sunny day),respectively.Predictably,the water loss in the two lines was not significantly different at 27 °C, relative humidity of 95%, and light intensity of 10,000 lx (Fig.S4A), whereas the dlr1 mutant lost water more rapidly than WT at the more extreme conditions of 37 °C, relative humidity 65%, and 20,000 lx (Fig.S4B).The stomatal density in dlr1 mutant and ko-1/2 leaves were also significantly decreased compared with the WT and COM(Fig.6B, F).

We measured stomatal conductance,transpiration rate,and net photosynthesis rate in the WT, dlr1 mutant, ko-1/2, and COM leaves using a steady-state porometer to test whether the alternation of stomatal movement and density affected physiologic processes associated with water status and photosynthesis.Stomata conductance, transpiration rate, and net photosynthesis rate were significantly lower in the dlr1 mutant and ko-1/2 leaves than in WT leaves(Fig.6G–I).Taken together,these results showed that mutation of the OsPG1 allele leads to defects in stomatal closure and a decreased stomatal density.

3.7.The OsPG1 mutant lines have reduced leaf pectinase activity and increased demethylesterified HGs in guard cells

Fig.5.Phenotypic analysis of roots.(A,B)Root phenotypes of WT,dlr1 mutant,ko-1/2,and COM plants at:10 d after sowing(DAS10,seedling stage)(A);tillering(B).Scale bars, 5 cm.(C) Root lengths at DAS5.(D) Numbers of lateral roots per centimeter 2-cm below the seed.(E) Lateral root length (mean length of the three lateral roots 2-cm below the seed(n=10)).(F–J)Transverse sections of roots at the grain filling stage.Scale bars,50 μm.(K–O)Propidium iodide staining of roots at the grain filling stage.Scale bars, 10 μm.Each measurement was based on 10 random plants (n = 10).Data are presented as means ± SD.

To explore the mechanism by which OsPG1 regulates the stomata density and closure,we speculated that the Q400R mutation in the OsPG1 protein might influence its enzyme activity.According to the InterPro database(https://www.ebi.ac.uk/interpro/)we confirmed that the point mutation was located in the Pectin_lyase_-fold domain of the OsPG1 protein (Fig.S5A).Online tool AlphaFold v2.3.2 (https://colab.research.google.com/github/deepmind/alphafold/blob/main/notebooks/AlphaFold.ipynb) was used to predict the protein structures of OsPG1 and OsPG1Q400R,respectively.The predicted structures between the 108th and 488th amino acids had very high model confidence (Fig.S5B), and the Q400R mutation did indeed alter OsPG1 protein structure(Fig.S5C).Further measurements were made of pectinase activity(including the activities of protopectinases,pectin methylesterases(PMEs), pectin lyases (PLs) and PGs (Fig.7A)) in the leaves of WT,dlr1 mutant, ko-1/2, and COM plants.The pectinase activities of dlr1 mutant and ko-1/2 leaves were significantly decreased in comparison to those of WT and COM leaves (Fig.7B).Because demethylesterified HGs are considered to be the substrate of PGs[35], we performed immunolabeling in longitudinal sections of WT,dlr1 mutant,ko-1/2,and COM leaves using JIM5 antibody that recognizes demethylesterified HGs [36].Antibody labeling was more intense in guard cells of the dlr1 mutant and ko-1/2 than WT and COM (Fig.7C).

Fig.6.Analysis of stomatal aperture and physiology.All tests were performed on lines WT,dlr1 mutant,ko-1/2,and COM.(A,B)Scanning electron microscope images of three types of stomatal aperture(A)and leaves(B).Yellow arrows indicate the stomata.Scale bars,1 mm.(C–E)Percentage of each of three levels of stomatal opening(40 stomata for each of replications) in the morning (C), noon (D), and evening (E) of a sunny day.(F) Statistical analysis of stomatal numbers per 24.26 mm2.Each measurement was repeated 10 times(n=10).(G–I)Measurements of leaf stomatal conductance(G),transpiration rate(H),and photosynthetic rate(I).Each measurement was repeated three times (n = 3).Data are means ± SD.*, P < 0.05, two-tail Student’s t-tests.

4.Discussion

4.1.The A1400G single base mutation in OsPG1 affects in vivo water homeostasis, thereby causing the dynamic leaf rolling phenotype

Plants adjust their architecture in adapting to changing environmental conditions[39,40].Leaf rolling is a typical adaptive trait in plants in response to temporary water shortage [4], and rolling or unfolding is controlled by the water content and size of bulliform cells [41].Excessive water loss leads to leaf tip withering and yellowing, and even wilting of the entire leaf.Identification of genes underlying dynamic plant architecture contributes to understanding the adaptability of plants to environmental change.Here, we identified the OsPG1/PHOTO-SENSITIVE LEAF ROLLING 1(PSL1) gene encoding a polygalacturonase [18,29] through the dlr1 mutant with dynamic leaf rolling(Fig.1).Conditional leaf rolling in the psl1 mutant varied with light intensity and humidity[18], whereas the leaf rolling phenotype in the dlr1 mutant appeared occurred under relatively extreme conditions, like the noon of a sunny day with relatively high light intensity and low humidity(Figs.1 and S1),indicating that the leaf rolling phenotype of the dlr1 mutant was associated with light and humidity.Further experiments demonstrated that leaf rolling in dlr1 mutant was caused by in vivo water deficiency, supported by measurements of water content and water spray treatments (Fig.2).The leaf tip necrosis in the dlr1 mutant (Fig.1K, L) was also a characteristic of leaves lacking water.The leaf rolling caused by water deficiency did not conflict with the light and humidity-induced leaf rolling,because the light and humidity were most likely to affect in vivo water content.The phenotype of the psl1 mutant was caused by the OsPG1 mutation involving a 260 bp deletion in the first exon,leading to a frameshift mutation and premature termination and thus a mutant OsPG1 protein containing the correct first ～120 amino acids [18].The ko-1/2 mutants also contained frameshift mutations and premature termination of the OsPG1 gene(Fig.S3C).However, the dlr1 mutant harbored an A1400G single base mutation in the OsPG1 gene,which resulted in a Q400R mutation in the OsPG1 protein, probably affecting the spatial structure of the protein (Fig.S5).Further experiments confirmed that the Q400R mutation resulted in reduced pectin activity in the dlr1 mutant (Fig.7B).Our findings collectively demonstrate that the Q400 is essential for the function of the OsPG1 protein and the Q400R mutation affects in vivo water homeostasis, thus causing the dynamic leaf rolling phenotype.

4.2.Mutation of OsPG1 probably does not affect water absorption by roots

Water uptake by plants mainly depends on roots,thus playing a central role in maintaining water homeostasis in a variable environment.Water transport occurs in interconnected tracheary elements (TEs) in the xylem [34,42,43].Abnormal root development affects water uptake and transport in plants.Root length and lateral root number in the OsPG1 mutant lines were not different from the WT(Fig.5A–E),although the OsPG1 gene was expressed in root tissues (Fig.4).Although OsPG1/PSL1 regulates cell wall integrity[18,29],the roots of the mutant lines showed no obvious difference in xylem formation and TE connections compared to the WT and COM (Fig.5F–O).These observations suggested that mutation of OsPG1 probably did not affect water absorption by the roots, but other influences of the mutation on root development could not be ruled out.

4.3.OsPG1 regulates stomatal closure and density to maintain water homeostasis

Stomata function as exit points or gateways linking the intercellular gas spaces to the surrounding atmosphere for photosynthetic gas exchange and transpiration[44,45].Stomata close and open to optimize gas exchange and water vapor,and closure reduces water loss in response to dehydration [45,46].OsPG1 gene mutants exhibited defective stomatal behavior (Fig.6D, E) that could lead to excessive water loss, partly supported by rapid water loss from dlr1 mutant detached leaves under relatively extreme conditions(Fig.S4).Stomatal conductance is determined by stomatal size and density and is positively correlated with transpiration[47,48].Despite the defects in stomatal closure in the mutant lines(Fig.6D,E),decreased stomatal density(Fig.6F)was most likely to be the main cause of decreased stomatal conductance(Fig.6G)and reduced transpiration (Fig.6H).Decreased stomatal conductance could also be due to reduced net photosynthesis rate in the mutants (Fig.6I).

Stomatal movement depends on guard cell shape which is controlled by turgor pressure[49].The degree of change in guard cells is limited by the mechanical properties of the cell wall[50].Pectins are proposed to be the main determinants of flexibility and elasticity in guard cell walls [51,52], which are composed of HG,rhamnogalacturonan-I(RG-I),rhamnogalacturonan-II(RG-II),xylogalacturonan, and apiogalacturonan [35,37].HG, representing up to 65% of the pectin in plant cell walls, is initially synthesized in a highly methylesterified form and can be demethylesterified by pectin methylesterases (PMEs); the demethylesterified HG can either form Ca2+-cross-linked pectin gels,resulting in cell wall stiffness, or be degraded to galacturonic acids by the action of pectin lyases (PLs) or PGs [37,53].The demethylesterified HG status has dramatic impacts on guard cell wall mechanics.A pectin methylesterase gene PME6 In Arabidopsis was highly expressed in guard cells, and its mutation led to guard cells being enriched in methylesterified HGs but having a decreased dynamic range of stomatal movements[50].A polygalacturonase(PG)gene POLYGALACTURONASE INVOLVED IN EXPANSION3 (PGX3) was expressed in guard cells and its mutation resulted in abundant demethylesterified HGs in guard cell walls,and limited stomatal movements[54].However, the two studies showed that mutants of both PME5 and PGX3 exhibited normal stomatal density.Here, we found that OsPG1 was highly expressed in stomata (Fig.4L), and its mutation led to defective stomatal closure (Fig.6D, E), and unlike PME5 and PGX3 in the Arabidopsis mutants,also resulted in decreased stomatal density.The OsPG1 mutant lines had decreased pectinase activity in the leaves and increased demethylesterified HGs in the guard cells (Fig.7), suggesting a role of OsPG1 in modulating demethylesterified HG abundance in guard cells in controlling the stomatal behavior.Plants growing in a changing environment constantly modulate stomatal activity by adjusting size and density to coordinate water loss and gas exchange [55,56].Therefore,the most plausible interpretation of our observations was that defective stomatal behavior in the OsPG1 mutant lines might lead to a loss of control of stomatal density,and the cost of such passive adjustment was decreased stomatal conductance and transpiration(Fig.6G, H), eventually affecting the water status of the whole plant.Further physiological experiments are needed to uncover the underlying mechanisms.Taken together, OsPG1 regulates demethylesterified HGs abundance in guard cells to modulate stomatal closure and density,thus maintaining water homeostasis.

Fig.7.Analysis of pectinase activity and immunolabeling.(A)Main components of pectin and enzymes involved in its degradation[35,37,38].(B)Measurements of pectinase activity of WT, dlr1 mutant, ko-1/2, and COM leaves.Pectinase activity represents the total activity of propectinases, pectin methylesterases, pectin lyases, and polygalacturonases.(C) Immunolabeling using the JIM5 antibody (recognizes demethylesterified HG) in guard cell walls of WT, dlr1 mutant, ko-1/2, and COM leaves.PMEs,pectin methylesterases;PLs,pectin lyases;PGs,polygalacturonases.Each measurement was repeated three times(n=3).Data are presented as means±SD.**,P<0.01,twotail Student’s t-test.Scale bars, 10 μm.

CRediT authorship contribution statement

Qinwen Zou:Project administration,Investigation,Formal analysis, Writing – original draft, Funding acquisition, Writing –review&editing.Ranran Tu:Project administration,Investigation,Formal analysis, Writing – review & editing.Jiajun Wu:Project administration, Investigation, Formal analysis.Tingting Huang:Investigation, Formal analysis.Zhihao Sun:Investigation, Formal analysis.Zheyan Ruan:Investigation, Formal analysis.Hongyu Cao:Investigation, Formal analysis.Shihui Yang:Investigation,Formal analysis.Xihong Shen:Resources,Writing–review&editing.Guanghua He:Conceptualization, Supervision, Resources,Writing–review&editing,Funding acquisition.Hong Wang:Conceptualization, Supervision, Resources, Writing – original draft,Writing – review & editing, Funding acquisition.

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 Postgraduate Research Innovation Project of Chongqing(CYS23217),Chongqing Modern Agricultural Industry Technology System (CQMAITS202301), the Science Fund for Creative Research Groups of the Natural Science Foundation of Chongqing, China (cstc2021jcyj-cxttX0004), and Natural Science Foundation of Chongqing (2023NSCQ-BHX0281).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2023.12.007.