The miR5810/OsMRLP6 regulatory module affects rice seedling photosynthesis

Weiwei Gao, Mingkang Li, Huaping Cheng, Kuaifei Xia*, Mingyong Zhang*

a State Key Laboratory of Plant Diversity and Specialty Crops & Guangdong Provincial Key Laboratory of Applied Botany, South China Botanical Garden, Chinese Academy of Sciences, Guangzhou 510650, Guangdong, China

b State Key Lab for Conservation and Utilization of Subtropical Agro-bioresources, College of Agriculture, University of Guangxi, Nanning 530004, Guangxi, China

c University of Chinese Academy of Sciences, Beijing 101408, China

d South China National Botanical Garden, Guangzhou 510650, Guangdong, China

Keywords:miR5810 OsNF-YB7 OsMRLP6 Photosynthesis Rice

ABSTRACT Photosynthesis affects crop growth and yield.The roles of microRNAs(miRNAs)in photosynthesis are little known.In the present study, the role of the OsNF-YB7–OsMIR5810–OsMRLP6 regulatory module in photosynthesis was investigated.The malectin-like protein gene OsMRLP6 was identified as a target gene of osa-miR5810 (miR5810).Overexpression in rice of miR5810 or down-expression of OsMRLP6 resulted in reduced expression of genes involved in chloroplast development and photosynthesis and decreased net photosynthetic rate, finally leading to lower shoot biomass and grain yield.Down-expression of miR5810 and overexpression of OsMRLP6 showed the opposite effect.Overexpression of transcription factor OsNF-YB7 elevated expression of OsMIR5810 in rice seedlings by binding to its promoter.The OsNFYB7–OsMIR5810–OsMRLP6 regulatory module affects photosynthesis to mediate growth and grain yield.

1.Introduction

Increasing rice(Oryza sativa L.)yield is a target of rice breeding.Improving photosynthetic efficiency has the potential to increase crop yields and reduce food shortages [1,2].Most studies aimed at increasing plant photosynthetic capacity of plants have focused on improving a particular step or property of photosynthesis [3].Identification of photosynthetic regulators that simultaneously regulate expression of many genes is another approach to improving photosynthesis[4].Previous studies have identified genes that directly or indirectly mediate photosynthesis,in particular nuclear genes [5].The transcription factor HYR (HIGHER YIELD RICE)directly activates genes involved in photosynthetic carbon metabolism to enhance photosynthetic rate in rice [6].Overexpression in rice of the maize transcription factor mEmBP-1 increased photosynthesis,biomass,and grain yield[3].The LEC1 transcription factor activated dozens of chloroplast biogenesis and photosynthesisrelated genes in Arabidopsis and soybean embryos [7,8].

Plant microRNAs (miRNAs) function in development and in response to biotic and abiotic stresses [9,10].In a transcriptomic survey [11], about 20% of pineapple miRNAs showed expression patterns with circadian oscillation, with several possibly targeting genes involved in crassulacean acid metabolism photosynthesis.

Arabidopsis miR408 mediates the HY5–SPL7 (SQUAMOSA PROMOTER BINDING PROTEIN-LIKE7) network by coordinating responses to light and copper [12].The dark-induced senescence plantacyanin-SAG14 (senescence-associated gene 14) module was repressed by miR408 in Populus [13].Rice miR408 downregulated its target gene OsUCL8 (UCLACYANIN-LIKE PROTEIN8)and positively regulated grain yield by regulating plastocyanin content and photosynthesis [14,15].Overexpression of miR398b increased salt sensitivity by regulating the antioxidant system and photosynthesis in tomato [16].High light stress induced expression changes of some miRNAs in Arabidopsis, and miRNAs may be regulated by reactive oxygen species (ROS) signaling[17].Populus miR156 can affect leaf composition change, and resulted in altering photosynthetic traits during vegetative phase change [18].

Malectin/malectin-like domain proteins are a subfamily of lectins that regulate many biological functions in plants[19].According to whether they contain a kinase domain, the malectin/malectin-like domain proteins are divided into MRLPs (malectin/malectin-like proteins) and MRLKs (malectin/malectin-like receptor-like kinases).The rice genome contains genes encoding 84 malectin/malectin-like domain proteins: 67 OsMRLKs and 17 OsMRLPs [19].Well-described MRLKs and MRLPs in plants are Catharanthus roseus receptor-like kinase-like proteins (CrRLK1Ls),and FERONIAs (FERs) are the best characterized CrRLK1L family proteins[20].FERs function in integrating hormones and stress signals with regulation of plant cell development and growth[21,22].FER was also reported to mediate the stability of 14-3-3 proteins in regulating plant carbon and nitrogen responses and growth via the RALF1 (rapid alkalinization factor peptide)/FER–ATL6 (Arabidopsis Tóxicos en Levadura 6)pathway in Arabidopsis[23].CrRLK1Ls have been reported to function in multiple biological roles from cellwall integrity to immunity, and most reported CrRLK1Ls contain a kinase domain [20].Most MRLP proteins, such as OsCBM1(OsMRLP15), lack a kinase domain but contain a malectin-like domain [24].OsCBM1 is involved in multiple signaling regulatory mechanisms, participating in NADPH oxidase-mediated ROS production by interacting with OsRacGEF1 and OsRbohA, resulting in increased tolerance of rice to drought stress[24].Only a few MRLPs have been identified in rice.OsMRLP6 is highly expressed in all tissues excluding mature embryo, endosperm, pre-emergence inflorescence, and young leaves [19].

In the present study, we aimed to explore effect of the OsNFYB7–OsMIR5810–OsMRLP6 regulatory module on rice seedling photosynthesis.

2.Materials and methods

2.1.Plant growth conditions

Transgenic rice was generated from Zhonghua 11 (ZH11), an Oryza sativa L.japonica cultivar.Rice seeds were surfacesterilized with 70% (v/v) ethanol for 30 s and 5% NaClO (w/v) for 40 min, then rinsed with water several times and germinated in darkness at 28°C for 3 d.The plants were grown in a closed paddy field of Guangzhou,Guangdong,China(23°10′47′′N,113°21′6′′E)or liquid cultured as previously described [25].

2.2.Vector construction, rice transformation and target gene prediction

To construct the overexpressing miR5810(miR5810-ox)vector,a 24-base-pair (bp) artificial miR5810 was cloned and inserted in pCAMBIA1301 vector (https://www.cambia.org) as previously described process [25].To construct down-regulated miR5810(STTM5810) vector, the Short Tandem Target Mimic method(STTM)was used as previously described with the pXU1301 vector[26].To construct the overexpressing OsMRLP6(MRLP6-ox)vector,the full-length coding sequence (CDS) of OsMRLP6 was amplified from leaves of ZH11 and cloned between the HindIII and BamHI sites downstream of the Ubi-1 promoter in the pXU1301 vector,thus OsMRLP6 CDS is fused with six hemagglutinin (HA) tag and one green fluorescent protein (GFP) tag.To construct OsMRLP6-RNAi vector, two fragments of the OsMRLP6 cDNA (342 bp) were amplified from ZH11 and inserted downstream of the Ubi-1 promoter in the rice RNA interference (RNAi) vector pTCK303 [27].To construct OsMIR5810pro:GUS and OsMRLP6pro:GUS vectors, a

～2-kb promoter located upstream of OsMIR5810 stem-loop or OsMRLP6 CDS was cloned and inserted into pCAMBIA1301 vector,respectively.For subcellular localization of OsMRLP6, the CDS of OsMRLP6 was cloned into the pUC18 vector to produce the 35S:OsMRLP6-GFP vector.These constructs were introduced into Agrobacterium tumefaciens strain EHA105 to transform ZH11 as previously described[25].All primers and genes used in this study are described in Table S1.The psRNATarget database(https://plantgrn.noble.org/psRNATarget/) was used to identify the target gene of miR5810.

2.3.RNA extraction, RT-qPCR, 5′RLM-RACE, and RNA-seq

The extraction of RNA(small RNA and total RNA),and quantitative real-time PCR(RT-qPCR)amplification were performed as previously described[25].5′-RNA ligase-mediated rapid amplification of cDNA ends (5′RLM-RACE) of total RNA from tillering-stage leaf sheaths of ZH11 and miR5810-ox was performed to identify miRNA cleavage sites on their target genes as previously described[25].Four-week-old miR5810-ox homozygotes and ZH11 seedlings grown in Hoagland solution at 25–28 °C under a 14 h/10 h light regime were sent to Novogene(Shanghai,China)for RNA sequencing (RNA-seq) [25].The transcripts were searched against the Swiss-Prot protein database (https://www.uniprot.org/) with BLASTX and the National Center for Biotechnology Information(NCBI) refseq RNA database (https://www.ncbi.nlm.nih.gov) with BLASTN using an E-value cut-off of 0.00000001.Gene ontology(GO) functional annotation was performed at the PANTHER database (https://www.pantherdb.org/).Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis was conducted at the online KEGG Automatic Annotation Server 2.1 (https://www.genome.jp/kaas-bin/kaas_main).

2.4.GUS assays and subcellular localization

Histochemical GUS (β-glucuronidase) staining and subcellular localization were performed as previously described [25,28], and were photographed under a light microscope (Leica DVM6, Leica Microsystems, Wetzlar, Germany) and a confocal laser scanning microscope (Leica SP8 STED 3X, Leica Microsystems), respectively.For subcellular localization of OsMRLP6,a 35S:OsMRLP6-GFP fusion vector was transiently expressed in ZH11 protoplasts, and before imaging, the protoplasts were dyed with 2 μmol L-1FM4-64 for 5 min to show the plasma membrane.

2.5.Measurement of photosynthetic traits

Total soluble sugar was measured by anthrone colorimetry[29,30] with modification.Leaf samples were dried at 80 °C for 72 h, and ground to powder.The powder sample (0.1 g) was extracted three times with 8 mL of 80% ethanol at 80 °C for 30 min.The supernatant was filtered and brought to 25 mL with 80%ethanol.The extract(1 mL)was incubated with 5 mL anthrone reagent at 95°C for 15 min,and then the reaction was stopped on ice and absorbance was measured at 620 nm.Total chlorophyll content was measured as described [31] with minor modification.Fresh leaves were harvested and a 0.15 g sample was used to extract chlorophyll with 20 mL 95%ethanol for 24 h,and the supernatant was collected.Contents of chlorophyll a and chlorophyll b were determined by measuring absorbance at 663 and 645 nm,and total chlorophyll content was then calculated as previously described [32].Net photosynthetic rates (Pn) of flag leaves were measured at tillering stage with a portable photosynthesis system LI-6400(LI-COR,Lincoln,NE,USA).Measurements were conducted from 10:00 to12:30 in the morning.Pnwere constructed at the indicated optimum temperature in a saturating chamber with 400 μL L-1CO2under light densities of 1500 μmol photons m-2s-1[33,34].Maximum quantum use efficiency of PSII (Fv/Fm) of flag leaves was measured with PAM-2100 fluorometer(Walz,Freistaat Bayern,Germany).Leaves were clamped with a leaf clip.Light was turned on after 15 min, and the probe was placed in the leaf clip.Fv/Fmwas read when Fvreached 200–400.Each plant was measured five times.

2.6.Transmission and scanning electron microscopy

Transmission electron microscopy(TEM)and scanning electron microscopy (SEM) were used to observe rice leaf structure at late tillering stage and mature endosperm, respectively.TEM and SEM preparation followed previous descriptions [25,35].For TEM,a series of 80–90 nm sections of leaves were cut with an Ultracut E ultramicrotome (Leica, Wetzlar, Germany) and attached to a formvar-coated copper grid (Electron Microscopy Sciences, Tokyo,Japan).Sections were viewed with a Jeol JEM-1010 TEM (Tokyo,Japan)at 15 kV.For SEM,mature seeds were transversely sectioned with a razor blade.The samples were dehydrated with a freeze dryer(JFD-310;Hitachi,Tokyo,Japan),and coated with gold palladium in six 30 s bursts using an ion sputterer (JEE-420, Hitachi).Then the samples were imaged with a JSM-6360LV SEM (Hitachi).

2.7.EMSA and dual-luciferase assay

Eelectrophoretic mobility shift assay (EMSA) and the dualluciferase assay were performed as previously described [35,36].For EMSA,the CDS of OsNF-YB7 was cloned into the pGEX4T-1 vector to generate a GST-YB7 fusion protein, and transformed into Escherichia coli strain BL21.GST-YB7 protein was induced with 0.5 mmol L-1isopropyl β-D-thiogalactoside at 16 °C for 12 h and purified using Glutathione Sepharose 4B (GE17-0756-011, MilliporeSigma, MO, USA) according to the manufacturer’s protocol.Then the complementary single-stranded oligonucleotides derived from 42 bp of the CCAAT-box and ABRE motify regions of the OsMIR5810 promoter were synthesized as DNA probes with biotin.The probes were synthesized by Invitrogen (Shanghai, China).EMSA was performed using a LightShift Chemiluminescent EMSA Kit (No.20148, Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer’s protocol.For the dual-luciferase assay in Nicotiana benthamiana leaves, the 2-kb promoter sequence of OsMIR5810 was amplified from rice and inserted into pGreenII0800-LUC vector [36] as the luciferase reporter construct(35S:REN–OsMIR5810pro:LUC).The CDS of OsNF-YB7 was cloned into the pCAMBIA1300-GFP vector as the effector construct (35S:-OsNF-YB7-GFP).The effector construct and the reporter construct were co-transformed into A.tumefaciens strain GV3101 and then infiltrated into 3-week-old N.benthamiana leaves by Agrobacterium injection.The activities of firefly luciferase (LUC) and Renilla luciferase (REN) were measured using a Dual-Luciferase Reporter Assay System (E2920, Promega, Madison, WI, USA).Relative luciferase activity was calculated as the ratio of LUC to REN(LUC/REN).

2.8.Data analysis

One-way ANOVA analysis with the least significant difference test was conducted with SPSS 19.0 software (IBM, Armonk, NY,USA).

3.Results

3.1.OsMRLP6 is a target gene of miR5810

A total of 29 genes were searched to be putative targets of miR5810 (Table S2).To further identify the target gene of miR5810 (Fig.1), a total of 36 miR5810-overexpressing rice lines(miR5810-ox) and 56 miR5810 down-expressing rice lines(STTM5810) were constructed.Five lines for each transgenic construct were randomly selected as experimental materials(Fig.S1A, B).Among these putative targets, the malectin-like protein gene OsMRLP6 (LOC_Os04g49690) was substantially downregulated in miR5810-ox plants and up-regulated in STTM5810 plants,compared with other targets(Figs.1A,S2).By RNA-seq,only the expression of OsMRLP6 was changed in miR5810-ox plants,compared with that in ZH11 (Tables S3,S4).These results showed that miR5810 could change the expression of OsMRLP6 and that the expression changes were opposite in miR5810-ox and STTM5810 plants, suggesting that OsMRLP6 might be a target gene of miR5810.To determine whether OsMRLP6 could be regulated by miR5810 in vivo, we further performed 5′-RLM-RACE to identify the miR5810-directed cleavage site.Sixteen clones of 5′-RLMRACE were sequenced,and first sequencing reads of 16 clones were at 2165 bp of OsMRLP6 mRNA, which is 40 bp downstream of the miR5810/OsMRLP6 mRNA complementary site (Fig.1B).

To confirm that the expression of OsMRLP6 could be repressed by miR5810, a rapid transient assay in N.benthamiana was performed (Fig.1C–F).We constructed vectors expressing GFP fused with the putative target site of OsMRLP6 at its 5′-terminus (35S:OsMRLP6-GFP) (Fig.1C) or with mutated target site (35S:OsMRLP6m-GFP) that abolished recognition by miR5810 (Fig.1D).The sequence of pre-miR5810 (miR5810 precursor sequence) was obtained from miRBase and cloned into the pBI121 vector to produce a miR5810-overexpression vector(35S:pre-miR5810).A.tumefaciens harboring plasmid 35S:OsMRLP6-GFP and 35S:pre-miR5810

were co-transformed into N.benthamiana leaves by infiltration,while co-transformation of 35S:OsMRLP6m-GFP and 35S:premiR5810 served as negative controls.The GFP intensity and protein concentration expressed from 35S:OsMRLP6-GFP gradually decreased commensurately with increasing 35S:pre-miR5810 concentration (OD600 nm = 0.0–0.9) in the cells of N.benthamiana leaves (Fig.1E, F).By contrast, the GFP intensity and protein concentration expressed from 35S:OsMRLP6m-GFP was unchanged when it was co-expressed with 35S:pre-miR5810 (Fig.1E,F).These results indicate that miR5810 can repress expression of OsMRLP6.

3.2.Expression patterns of OsMRLP6 and miR5810

We then investigated the spatial expression patterns of miR5810 and OsMRLP6 in various organs of ZH11 (Fig.S3).The expression levels of miR5810 and OsMRLP6 showed a strong negative correlation in the development of leaves, leaf sheaths, stems,panicles, and developing seeds (Fig.S3A).This expression pattern of OsMRLP6 was similar to the data (Fig.S3B, C) from the Rice eFP database (https://www.bar.utoronto.ca/eplant_rice).GUS staining analysis of OsMIR5810 and OsMRLP6 promoter driving GUS transgenic rice plants showed that OsMIR5810 and OsMRLP6 was expressed in all organs, and the expression patterns of OsMIR5810 were also negatively correlated with those of OsMRLP6 in mature leaves, seedling roots, stems, and ligules (Fig.S3D).We concluded that miR5810 can target OsMRLP6 mRNA, thereby suppressing expression of OsMRLP6.We then generated OsMRLP6 transgenic rice plants,including OsMRLP6-overexpressing (MRLP6-ox) and down-expressing (MRLP6-RNAi) lines (Fig.S1C, D).

OsMRLP6 has been predicted to be a repeat receptor-like protein [19] that contains a malectin-like domain and a transmembrane domain (Fig.2A).For subcellular localization of OsMRLP6,its CDS was fused with GFP.The vectors 35S:OsMRLP6-GFP and 35S:GFP (as a control) were introduced separately into rice protoplasts by PEG-mediated transformation.The fluorescence signal of OsMRLP6-GFP was strongly co-localized with FM4-64.However,the signal of GFP alone as the control was detected throughout cells, indicating that OsMRLP6 is localized mainly in the plasma membrane (Fig.2B).

3.3.miR5810 and OsMRLP6 affect photosynthesis in rice seedlings

Fig.1.OsMRLP6 is a target gene of miR5810.(A) Expression change of OsMRLP6 in transgenic rice lines with miR5810 overexpression (miR5810-ox) and down-expression(STTM5810).e-EF-1a was used as reference gene.Error bars indicate SD with triplicates.**,P<0.01 by Student’s t-test.(B)Cleavage site identification of OsMRLP6 mRNA.The arrow indicates the cleavage site identified in ZH11 seedlings by 5′-RLM-RACE.(C–F)Transient assay of OsMRLP6 expression repressed by miR5810 in N.benthamiana leaves.The 24-bp target fragment OsMRLP6 (C) and the 24-bp mutated target fragment OsMRLP6m (D) were fused in front of the GFP gene and introduced into Agrobacterium.The mutated bases are shown in red(D).Then the fluorescence intensity(E)and Western blotting(F)of GFP were read after co-transformation of pre-miR5810 with 35S:OsMRLP6-GFP or 35S:OsMRLP6m-GFP into N.benthamiana leaves, respectively.OD600 in (F) indicates the concentration of Agrobacterium with 35S:pre-miR5810, 35S:OsMRLP6-GFP or 35S:OsMRLP6m-GFP.Scale bar, 25 μm in (E), and the immunoprecipitated protein was detected using an anti-GFP and anti-actin antibody (F).

To investigate how miR5810 and OsMRLP6 play roles, RNA-seq of the leaves of ZH11 and miR5810-ox was performed.Of the transcripts,381 were down-regulated and 578 up-regulated in T4 lines of miR5810-ox (Tables S3, S4).Ten of the genes with altered expression identified by RNA-seq were verified by RT-qPCR(Fig.S4).Among the down-regulated genes from RNA-seq, most were involved in chloroplast development and photosynthesis(Figs.3A, S5).Fourteen genes were involved in photosynthesis(Table S5):seven antenna protein genes(Lhca2,Lhca4,Lhca5,Lhcb1,Lhcb3, Lhcb5, and Lhca6), four photosystem I and II (PSI and PSII)reaction center subunits (PsaN, PsaO, PsbP, and Psb27), and three genes involved in carbon fixation (PRK, ALDO, and FBP).RT-qPCR verified that five of these genes (Lhca2, Lhca3, Lhca5, PsbP, and PsaO) were down-regulated in miR5810-ox and MRLP6-RNAi(Fig.S6).The expression of three of those genes:OsLhca2, OsLhca4,and OsLhca, was up-regulated in MRLP6-ox and STTM5810 plants,compared with those in ZH11 (Fig.3B).

To clarify effect of miR5810 and OsMRLP6 on photosynthesis,photosynthetic parameters of the miR5810 and OsMRLP6 transgenic rice plants were measured (Fig.3C–E).Pnand (Fv/Fm) of flag leaves were measured in the transgenic rice seedlings (Fig.3C, D).miR5810-ox and MRLP6-RNAi plants exhibited lower Pnand Fv/Fmthan ZH11 plants,while STTM5810 and MRLP6-ox plants exhibited higher Pnand Fv/Fmthan ZH11 plants.We further analyzed the soluble sugar concentration of these transgenic plants because photosynthetic rate directly regulates carbon assimilation.Consistently,miR5810-ox and MRLP6-RNAi plants accumulated less soluble sugars in leaves than ZH11 plants (Fig.3E).We accordingly hypothesized that miR5810 and OsMRLP6 mediate photosynthesis in rice seedlings.

To investigate chloroplast development in the miR5810 and OsMRLP6 transgenic rice plants, the densities and ultrastructure of chloroplast in the top full leaves at the tillering stage were tested using TEM(Fig.4A,B).As in ZH11,the chloroplast structures in the STTM5810 and MRLP6-ox plants displayed well-developed membrane structure with dense thylakoids arranged in the grana, and the grana lamellae structures were very clear.However, the miR5810-ox and MRLP6-RNAi plants displayed abnormal chloroplast structures, including enlarged and deformed chloroplasts and indistinct grana lamellae; the grana spaces were also larger and non-cohesive (Fig.4A, B).Transcripts of many genes involved in chlorophyll biosynthesis and metabolism processes were down-regulated in miR5810-ox (Fig.S5; Table S3).Consistently,the total chlorophyll content of miR5810-ox and MRLP6-RNAi leaves was significantly lower than that of ZH11, whereas those of STTM5810 and MRLP6-ox were higher than that of ZH11, indicating that the changes in photosynthesis were accompanied by increased leaf chlorophyll (Fig.4C).

Fig.2.OsMRLP6 is localized in the plasma membrane.(A)Schematic structure of OsMRLP6 protein domains.SP, signal peptide.Tr,transmembrane domain.(B)Subcellular localization of OsMRLP6 in rice protoplasts.Vector construct 35S:OsMRLP6-GFP and 35S:GFP were transformed separately into rice protoplasts.Before imaging,the protoplasts were treated with FM4-64 (2 μmol L-1) for 5 min to show the plasma membrane.35S:GFP was used as a negative control.Scale bar, 5 μm.

Fig.3.miR5810 and OsMRLP6 affect photosynthesis of rice seedlings.(A) go enrichment analysis for the 29 down-regulated genes with the largest expression changes in mir5810-ox.(B) expression changes of genes involved in the photosynthetic electron transport chain in miR5810 and OsMRLP6 transgenic rice seedlings.(C,D) net photosynthetic rate (C), maximum quantum use efficiency (D), and soluble sugar content (E) of the mir5810 and osmrlp6 transgenic rice leaves.dw, dry weight.error bars indicate sd with triplicates; *, P < 0.05, **, P < 0.01 by Student’s t-test.

Starch is the main product of photosynthesis.Starch granules in the chloroplast of mesophyll cells were also further studied with TEM (Fig.S7).As in ZH11, the starch granules in the chloroplast of STTM5810 and MRLP6-ox plants appeared as normal oval shapes(Fig.S7).By contrast,the miR5810-ox and MRLP6-RNAi plants displayed more and smaller circular starch granules in the chloroplast(Fig.S7), which may affect the synthesis of starch in seeds.SEM(Fig.4D) revealed that the starch granules of the seed endosperm cells in ZH11 were polygonal with sharply defined edges and arranged in tightly compound granules; the starch granules in STTM5810 and MRLP6-ox plants were similar to that of ZH11.By contrast, the seed endosperm cells in miR5810-ox and MRLP6-RNAi plants showed loosely packed starch granules, arranged as spherical granules.

Fig.4.miR5810 and OsMRLP6 affect chloroplast development and grain starch accumulation in rice.(A,B)Transmission electron microscopy(TEM)analysis of chloroplast(A)and grana lamellae(B)of the miR5810 and OsMRLP6 transgenic rice.Scale bars,1 μm(A)and 500 nm(B).ch,chloroplast;gl,grana lamellae.At least six seedlings of each line were used for TEM.(C) Total chlorophyll content of flag leaves in miR5810 and OsMRLP6 transgenic rice.FW, fresh weight.Experiments were performed in three biological replicates with similar results.Values shown are means±SD of six independent seedlings.*,P<0.05,**,P<0.01 by Student’s t-test.(D)Scanning electron microscopy(SEM)of mature endosperms.Scale bar, 10 μm.CSG, compound starch granule; SSG, spherical starch granule.At least six seedlings of each line were used for SEM analysis.

3.4.miR5810 and OsMRLP6 showed opposite effects on rice growth and grain yield

To investigate whether miR5810 and OsMRLP6 affect rice growth and grain yield,the main traits were measured under normal conditions in a paddy field(Figs.5, S8).Compared with ZH11,miR5810-ox plants showed semi-dwarfing, greater tiller numbers and earlier-maturity phenotypes,and the opposite tendency of tiller number was observed in STTM5810 plants (Fig.S8).The plant height and tiller number of MRLP6-RNAi were slightly reduced relative to those of ZH11(Fig.S8).The shoot dry biomass,seed setting rate and grain yield of miR5810-ox and MRLP6-RNAi were lower,whereas the shoot dry biomass and grain yield of STTM5810 and OsMRLP6-ox were higher than those of ZH11 (Fig.5).Thus,miR5810 and OsMRLP6 play respectively negative and positive roles in rice growth and grain yield,.

3.5.Expression of OsMIR5810 is regulated by OsNF-YB7

A lot of the upstream regulators of OsMIR5810 were identified in the 2-kb upstream sequence of OsMIR5810 promoter (Table S6),including a CCAAT-box and an ABRE motif (Fig.S9).The nuclear transcription factor Y (NF-Y) of the rice LEC1 family is involved in regulating photosynthesis and chloroplast biogenesis, and can bind to CCAAT DNA sequences [37].We found that the rice OsNF-YB7 may bind to the CCAAT-box and ABRE motif of the OsMIR5810 promoter (Fig.S10).To further test whether OsNFYB7 could directly bind to the promoter of OsMIR5810, we performed EMSA with the purified GST–OsNF-YB7 recombinant protein (GST–YB7) and OsMIR5810 probes (Fig.S10).Based on the CCAAT-box and ABRE motif in OsMIR5810 promoter (Fig.S10), we designed two labeled probes (P1and P2) for EMSA (Fig.S10A;Table S6).OsNF-YB7 bound specifically to the probe region of the OsMIR5810 promoter, and EMSA binding was substantially weakened by non-labeled competitive probe in a dosage-dependent manner (Fig.S10).The protein–DNA binding was compromised when some of the nucleotides in the conserved CCAAT-box in the P1region were mutated,showing that the CCAAT-box acts as a core site for recognition by OsNF-YB7 (Fig.6A, B).To determine how OsNF-YB7 regulates OsMIR5810 expression, we tested the expression of pre-miR5810 (OsMIR5810) and the mature miR5810 in OsNF-YB7-overexpressing lines (YB7-ox) (Figs.6C, S11).The expression of pre-miR5810 was substantially up-regulated in YB7-ox rice plants (Fig.6C).The mature miR5810 also showed higher expression in YB7-ox(Fig.S11B).We subsequently conducted transient expression assays using a dual-luciferase (LUC) reporter to confirm the result in vivo.The dual luciferase reporter plasmids contain LUC driven by the OsMIR5810 promoter (MIR5810pro:LUC)and REN driven by the CaMV35S promoter as an internal control,while the 35S:OsNF-YB7 plasmid was used as effector.The coexpression of MIR5810pro:LUC and 35S:OsNF-YB7 significantly increased LUC reporter activity, in contrast to the effects of the GFP empty vector control(Fig.6D–F).Together,these results show that OsNF-YB7 functions as a transcriptional activator of OsMIR5810 expression by binding to the CCAAT-box and ABRE elements.

Fig.5.miR5810 and osmrlp6 affect biomass and grain yield of rice.(A–C)statistical analysis of shoot dry biomass(A),seed setting rate(B)and grain yield per plant(C)of the mir5810 and osmrlp6 transgenic rice lines grown in a controlled field at mature stage.each line was measured in at least 20 independent plants.error bars indicate sd with triplicates; *, P < 0.05, **, P < 0.01, ***, P < 0.001 by Student’s t-test.(D) Photos of rice panicle of the miR5810 and OsMRLP6 transgenic rice plants.Scale bar, 2 cm.

OsNF-YB7 had been found [7] to be expressed mainly in developing seeds (Fig.S12A, B), especially in embryos, and may also be expressed at a low level in young leaves(Fig.S12A).We measured the diurnal expression of OsNF-YB7 in shoots of 2-week old ZH11 rice seedlings.Expression of OsNF-YB7 in shoots of rice seedlings was higher during the day than at night, especially between 10:00 and 16:00, with the highest expression at 12:00 (Fig.S12C,D).But the target cycle threshold (CT) values exceeded 32 cycles,suggesting that expression of OsNF-YB7 is lower in leaves of seedlings.But this result implies that OsNF-YB7 may also be less highly expressed in some cells of young leaves.We inferred that OsNFYB7 may regulate the expression of OsMIR5810 to affect photosynthesis of rice seedling.

3.6.Genetic variations in OsMIR5810

Considering that miR5810 functions in regulating rice grain yield, we investigated the variation in the pre-miR5810 sequence and its promoter sequences among 5459 rice varieties using MBKbase (https://www.mbkbase.org/).There were two singlenucleotide polymorphisms (SNPs) and one insertion–deletion polymorphism (InDel) in the pre-miR5810 sequences of all rice varieties.Most of the rice varieties(90.3%,4930/5459)were classified into two haplotypes by the two SNPs(HAP1 with C-T SNPs and HAP2 A–C with SNPs in the pre-miR5810 sequences, respectively)(Fig.S13A).The majority of indica varieties(78.3%,1990/2542)carried the HAP1 haplotype (Fig.S13C).By contrast, the HAP2 haplotype was observed mainly in Temperate japonica (59.8%,1428/2388) and Tropical japonica (17.6%, 420/2388) varieties.In the promoter of OsMIR5810, ten SNPs were identified in all examined accessions.Most of the accessions (72.3%, 3947/5459) were classified into two haplotypes, which contained four SNPs and were represented by Nipponbare (MIR5810NIP) and IR64(MIR5810IR), respectively (Fig.S13B).Similar to the haplotypes classified by the mature miR5810 sequence, the majority of indica varieties (78.3%, 1990/2542) were MIR5810IRhaplotypes, whereas only a small fraction of the Temperate japonica varieties (6.6%,169/2542) were classified as MIR5810IRhaplotypes.By contrast,the majority of Temperate japonica varieties (59.8%, 1428/2388)were classified as MIR5810NIP(Fig.S3D).HAP1 based on two SNPs in the miR5810 precursor sequences was well correlated with MIR5810IRhaplotypes (1520 varieties carried both HAP1 and MIR5810IR), whereas HAP2 was correlated with MIR5810NIPhaplotypes (2048 varieties carried both HAP2 and MIR5810NIP)(Fig.S13E).

We measured the expression of miR5810 and OsMRLP6 in varieties of HAP1 (TeQing, Jiangxisimiao, ASWINA 330, Q5, Zegu, and Guangluai4hao) and HAP2 (M202, WIR 911, Nipponbare, Lemont,and JinghuB).Expression of miR5810 in HAP1 varieties was lower than that in HAP2 varieties (Fig.7A).The expression of OsMRLP6 was opposite to that of miR5810 in HAP1 and HAP2 varieties(Fig.7B).The expression changes of genes involved in photosynthesis (Lhca4 and Lhca5) in HAP1 and HAP2 varieties were similar to that of OsMRLP6 (Fig.7C–D).These results imply that genetic variations of pre-miR5810 in HAP1 and HAP2 varieties may affect their photosynthesis.

4.Discussion

In this study, we found that OsNF-YB7–OsMIR5810–OsMRLP6 module affects photosynthesis in rice.Downregulation of miR5810 or overexpression of OsMRLP6 in rice may be further translated into improving biomass production and grain yield since their enhanced photosynthesis.However, overexpression of miR5810 or downregulation of OsMRLP6 resulted in opposite phenotypes (Figs.3, 4, 5, S8).Our findings suggest that miR5810 or OsMRLP6 can be used as genetic target genes to increasing photosynthetic capacity in rice.

Fig.6.OsNF-YB7 binds to the promoter of OsMIR5810 and activates its expression.(A) Schematic representation of CCAAT-box in the OsMIR5810 promoter.P1 represent probe positions for electrophoretic mobility shift assay (EMSA).(B) In vitro EMSA using CCAAT-box sequence from promoter of OsMIR5810 as probes.The P1 probe was a biotin-labeled fragment of the OsMIR5810 promoter and the competitor was an unlabeled competitive probe.The gradient indicates the increasing amount of competitor.The CCAAT-box sequence CCGTTG was substituted by CCGaaa (mutant probe1) and CaaaaG (mutant probe2) in the mutant probe.GST-YB7, fusion protein GST-OsNF-YB7; GST, negative control.(C) Expression levels of OsMIR5810 in OsNF-YB7 overexpressing rice (YB7-ox).U6 was used as miRNA reference,means±SD(n=3)are shown.*,P<0.05,**,P<0.01 by Student’s t-test.(D–F)In vivo luciferase assay of miR5810 expression activated by OsNF-YB7.(D) Schematic diagram of various constructs used in the transient transformation assay.35S:REN–MIR5810pro:LUC was constructed as reporter and 35S:YB7-GFP was constructed as effector.(E–F)OsNF-YB7 activates the transcription of OsMIR5810 in N.benthamiana.D-luciferin was used as the substrate of Luciferase in (E), free GFP was used as a negative control, and the expression of REN was used as an internal control.The LUC/REN ratio represents the relative activity of the OsMIR5810 promoter.Error bars indicate SD with biological triplicates (n=3),***, P<0.001 by Student’s t-test.

Here, miR5810 was identified as a negative regulator of rice grain yield via directly targeting OsMRLP6 (Figs.1, 5).Overexpression of miR5810 or down-regulation of OsMRLP6 in rice plants resulted in a decrease in shoot dry biomass, seed setting rate,and grain weight per plant, whereas down-regulation of miR5810 or overexpression of OsMRLP6 produced the opposite phenotypes(Fig.5).But inversely altered expression of miR5810 and OsMRLP6 did not always cause similar growth phenotypic changes in transgenic rice, such as in plant height and tillering number (Fig.S8).Thus,some other genes may also contribute to miR5810 regulating photosynthesis and grain yield.However, among these putative targets, only the expression of OsMRLP6 was changed in miR5810-ox plants.There were several cytochrome P450 genes among the differentially expressed genes,which may also regulate rice growth and development(Tables S3,S4).These results suggest that miR5810 may also target other genes in addition to the putative targets.Unlike MRLK proteins, which contain kinase domain,malectin/malectin-like and transmembrane domains [19],OsMRLP6 has only malectin-like and transmembrane domains,suggesting that its function may be more complex than that of MRLK proteins.The first described MRLK proteins in plants are FERONIA(FER)and THESEUS1(THE1)in Arabidopsis[38,39].Other MRLK proteins have been reported [40] to function in regulating plan growth and development.But few MRLP proteins have been studied,especially in rice.In a recent study[19],the spatial expression patterns of OsMRLPs in rice various developmental stages/tissues were characterized.OsMRLP genes displayed differing expression profiles among the tested tissues, OsMRLP4 and OsMRLP7 were highly expressed in all tested tissues and OsMRLP6 was also expressed in all tested tissues except embryo and 25 d after fertilization (DAF), endosperm at 25 DAF, pre-emergence inflorescence, and 20 days old leaves [19].These findings suggested that the function of OsMRLP6 may be repressed in these tissues.Our results suggest that OsMRLP6 is expressed in all tested tissues (Fig.S3), and acts in regulating grain yield (Fig.5), but the molecular mechanism underlying the OsMRLP6-regulated rice grain yield awaits investigation.

The miR5810/OsMRLP6 regulatory module is involved in photosynthesis.Photosynthetic rates and soluble sugar content were decreased in miR5810-ox and OsMRLP6-RNAi transgenic rice plants owing to the abnormal chloroplast structures, compared with ZH11(Figs.3C–E,4A,B).These results are consistent with the finding that genes involved in photosynthesis were down-regulated in miR5810 overexpression or OsMRLP6 down-regulation lines(Fig.3A, B; Table S5).Chlorophyll a/b binding proteins function in photosystems I and II and their disruption reduces photosynthetic efficiency [41,42].Down-regulation of Lhca2, Lhca4, Lhca5,PsbP, and PsaO in miR5810-ox and OsMRLP6-RNAi may result in a decrease in photosynthesis (Figs.3B, S6).Chloroplast biogenesis and development influence photosynthetic efficiency[43].Expression of many chloroplast development genes was changed in miR5810-ox (Fig.S5; Table S3).All these findings suggest that miR5810 and OsMRLP6 strongly influence photosynthesis.

Expression of OsMIR5810 was activated by transcription factor OsNF-YB7 in rice seedlings(Figs.6,S11).OsNF-YB7 had been found to be expressed mainly in the embryo and caryopsis at 5 DAF, and the osnf-yb7 mutant showed activation of many genes responsible for rice chlorophyll biogenesis in embryo at either 5 or 10 DAF[37,44].However, OsNF-YB7 (also known as OsLEC1/OsHAP3E) has been found [37] to repress chlorophyll biogenesis and photosynthesis in rice embryos.Mutation of OsLEC1 triggered embryo greening and promoted a series of genes involved in photosynthesis and photomorphogenesis [44,45].OsNF-YB7 is expressed at the adaxial side of the leafy primordium and the increased expression of OsHAP3E with OsNF-YB7 promoter caused abnormal leaf development [46].The Rice eFP database showed that OsNF-YB7 displayed highest expression in the seed, followed by the inflorescence and young leaves (Fig.S12A, B).Our results showed that OsNF-YB7 can directly bind to OsMIR5810 promoter (Fig.6).The promoter of OsNF-YB7 contains many light-responsive elements and the expression of OsNF-YB7 in seedling leaves could be induced at a low level by light (Fig.S12C, D; Table S7).OsNFYB7 mRNA was found to be accumulated mainly at primordium initiation [46], which may be one of the reasons for the lower expression in RT-qPCR detection (Fig.S12D).We infer that OsNFYB7 functions not only in the embryo but also in leaves to influence the photosynthetic capacity of rice by controlling the expression of OsMIR5810.OsMRLP6 displayed low expression in embryos [19],suggesting that the activated expression of OsMIR5810 in the embryo by OsNF-YB7 may down-regulate the expression of OsMRLP6 in the embryo.

We propose a working model (Fig.S14) for the roles of OsNFYB7-OsMIR5810-OsMRLP6 regulatory module in rice grain yield development by affecting seedling photosynthesis.Expression of OsMIR5810 is positively regulated by the transcription factor OsNF-YB7.When OsNF-YB7 accumulates, it binds to the CCAATbox or ABRE motif in the promoter of OsMIR5810 to activate its expression.The increased production of miR5810 then negatively regulates rice photosynthetic capacity by down-regulating the expression of OsMRLP6 and other targets,causing lower grain yield.

Fig.7.Expression of miR5810(A),OsMRLP6(B),OsLhca4(C),and OsLhca5(D)in HAP1 and HAP2 varieties.All HAP1 and HAP2 varieties were derived from 5459 rice accessions in the MBKbase database.e-EF-1a was used as reference gene.Error bars indicate SD with triplicates.*, P < 0.05 by Student’s t-test.

CRediT authorship contribution statement

Weiwei Gao:Conceptualization, Formal analysis, Investigation,Methodology, Writing-original draft-review & editing.Mingkang Li:Conceptualization,Formal analysis,Investigation,Methodology.Huaping Cheng:Investiagtion,Methodology.Kuaifei Xia:Conceptualization, Investigation, Methodology, Funding acquisition.Mingyong Zhang:Conceptualization, Resources, Project administration, Funding acquisition, Writing-review & editing,Supervision.

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

We thank Dr.Baixiao Niu (Yangzhou University) for providing YB7-ox transgenic rice plants.This study was supported by the Guangzhou Science and Technology Project (202102021003,2023B03J0742) and the National Natural Science Foundation of China (32171933).