LEAF TIP RUMPLED 1 Regulates Leaf Morphology and Salt Tolerance in Rice

Leaf morphology is one of the important traits related to ideal plant architecture and is an important factor determining rice stress resistance, which directly affects yield. Wax layers form a barrier to protect plants from different environmental stresses. However, the regulatory effect of wax synthesis genes on leaf morphology and salt tolerance is not well-understood. In this study, we identified a rice mutant, leaf tip rumpled 1 (ltr1), in a mutant library of the classic japonica variety Nipponbare. Phenotypic investigation of NPB and ltr1 suggested that ltr1 showed rumpled leaf with uneven distribution of bulliform cells and sclerenchyma cells, and disordered vascular bundles. A decrease in seed-setting rate in ltr1 led to decreased per-plant grain yield. Moreover, ltr1 was sensitive to salt stress, and LTR1 was strongly induced by salt stress. Map-based cloning of LTR1 showed that there was a 2-bp deletion in the eighth exon of LOC_Os02g40784 in ltr1, resulting in a frameshift mutation and early termination of transcription. Subsequently, the candidate gene was confirmed using complementation, overexpression, and knockout analysis of LOC_Os02g40784. Functional analysis of LTR1 showed that it was a wax synthesis gene and constitutively expressed in entire tissues with higher relative expression level in leaves and panicles. Moreover, overexpression of LTR1 enhanced yield in rice and LTR1 positively regulates salt stress by affecting water and ion homeostasis. These results lay a theoretical foundation for exploring the molecular mechanism of leaf morphogenesis and stress response, providing a new potential strategy for stress-tolerance breeding.


Introduction
Leaves are the main photosynthetic organ of plants. Leaf morphology affects the effective photosynthetic area, which affects accumulation of photosynthetic products and subsequent crop yield. In rice, numerous genes associated with leaf morphogenesis have been mined and cloned, such as SHALLOT-LIKE 1 (SLL1) [1], HOMEODOMAIN CONTAIN-ING PROTEIN4 (OsHB4) [2], SEMI-ROLLED LEAF1(SRL1) [3], Rice outermost cell-specific gene 5 (Roc5) [4], AGO1 homologs 1b (OsAGO1b) [5], Rice outermost cell-specific 8 (Roc8) [6], PHOTO-SENSITIVE LEAF ROLLING 1 (PSL1) [7]. These genes regulate leaf morphogenesis through complex interactions among plant hormone signaling pathways, transcription factors, and microRNAs [8,9]. In addition, leaf morphology is also affected by genes associated with ribosomes synthesis, DNA repair, cell cycle process, cuticle development, ion homeostasis, and microtubule arrangement [9]. However, these genes alone are not sufficient leaves, decreases in the seed setting rate and yield, and high sensitivity to salt stress. We also confirmed that LTR1 mediated regulatory activities of aquaporin and ion transporters result in altered water retention and ion homeostasis under salt stress. Hence, function analysis of LTR1 in leaf morphology and response to salt stress could provide theoretical foundation for molecular mechanism of leaf morphogenesis and salt response in rice and contribute to breeding efforts to develop salt-tolerant varieties with ideal leaf morphology.

Identification of the ltr1 Mutant
The ltr1 mutant was successfully obtained by EMS mutagenesis of the NPB. Phenotypic observation indicated that ltr1 exhibited abnormal leaf morphology with uneven distribution of bulliform cell on adaxial surface and sclerenchyma cells on abaxial surface and disordered vascular bundles (Figure 1a-c). The contents of chlorophyll a, chlorophyll b, and carotenoids were significantly higher in ltr1 than in NPB, with increases of 19.23%, 24.96%, and 17.83%, respectively (Figure 1d). The SPAD (soil and plant analyzer development) value of ltr1 was significantly higher than that of NPB ( Figure 1e). The quantum efficiency of photosystem II (Fv/Fm) and leaf water content of ltr1 were significantly lower than those of NPB, decreased by 8.69% and 5.26%, respectively (Figure 1f,g). These results showed that growth and development of ltr1 were seriously impaired. The abnormal leaf morphology of ltr1 was associated with lower light energy conversion efficiency of the PS II (Photosystem II) reaction center and poor leaf water retention.
focused on pollen development, panicle fertility, and drought resistance; there are few reports on the regulatory role of wax synthesis genes in leaf morphology. Here, we identified a rice mutant ltr1 with abnormal leaf morphology. This mutant was obtained by ethyl methanesulfonate (EMS) mutagenesis of Nipponbare and was used to isolate and analyze the function of the candidate gene LTR1 in regulating leaf morphology. We demonstrated that loss function of LTR1 led to abnormal development of bulliform cells, vascular bundles, and sclerenchyma cells, and to rumpled leaves, decreases in the seed setting rate and yield, and high sensitivity to salt stress. We also confirmed that LTR1 mediated regulatory activities of aquaporin and ion transporters result in altered water retention and ion homeostasis under salt stress. Hence, function analysis of LTR1 in leaf morphology and response to salt stress could provide theoretical foundation for molecular mechanism of leaf morphogenesis and salt response in rice and contribute to breeding efforts to develop salt-tolerant varieties with ideal leaf morphology.

Identification of the ltr1 Mutant
The ltr1 mutant was successfully obtained by EMS mutagenesis of the NPB. Phenotypic observation indicated that ltr1 exhibited abnormal leaf morphology with uneven distribution of bulliform cell on adaxial surface and sclerenchyma cells on abaxial surface and disordered vascular bundles (Figure 1a-c). The contents of chlorophyll a, chlorophyll b, and carotenoids were significantly higher in ltr1 than in NPB, with increases of 19.23%, 24.96%, and 17.83%, respectively (Figure 1d). The SPAD (soil and plant analyzer development) value of ltr1 was significantly higher than that of NPB ( Figure 1e). The quantum efficiency of photosystem II (Fv/Fm) and leaf water content of ltr1 were significantly lower than those of NPB, decreased by 8.69% and 5.26%, respectively (Figure 1f,g). These results showed that growth and development of ltr1 were seriously impaired. The abnormal leaf morphology of ltr1 was associated with lower light energy conversion efficiency of the PS II (Photosystem II) reaction center and poor leaf water retention.

Effect of LTR1 on Photosynthetic Efficiency and Seed Setting Rate
According to our results, the panicle length (Figure 2a,b), seed setting rate (Figure 2f), secondary branch numbers (Figure 2e), and grain yield per plant ( Figure 2h)were significantly lower in ltr1 plants than in the wild type, by 14.50%, 95.54%, 16.67%, and 84.49%, respectively. The effective panicle number was significantly higher for ltr1 than for NPB, with a 43.38% increase (Figure 2c). The primary branch numbers and 1000-grain weight showed no significant differences (Figure 2d,g). These results indicated that the decrease of yield per plant in ltr1 was caused by the extremely low seed-setting rate and showed that ltr1 had serious defects in leaf morphology and fertility. and high chlorophyll content is conducive to light absorption; the photosynthetic rate under saturated light intensity is mainly affected by the catalytic ability of the Rubisco enzyme, rather than the limitation of electron transfer rate in light reactions [38]. To explore whether the increased chlorophyll content and abnormal leaf morphology of ltr1 affect photosynthetic efficiency, we measured the photosynthetic efficiency of NPB and ltr1 in the field. Compared to NPB, the intercellular CO2 concentration of ltr1 was 4.91% higher, and the photosynthetic efficiency was 25.33% lower (Figure 2i,j). Although the photosynthetic pigment content of ltr1 increased, the photosynthetic efficiency did not. The reasons for the decrease of photosynthetic efficiency in ltr1 require further exploration.  branches, (f) seed setting rate, (g) 1000-grain weight, (h) grain yield per plant, (i) photosynthetic efficiency, and (j) intercellular CO 2 concentration of NPB and ltr1. Data are given as means ± SD. Asterisks indicate significant difference based on the Student's t-test: ** in the figure represents significant difference at p < 0.01 and ns in the figure represents there is no significant different at p < 0.05.
Photosynthesis is the sum of a series of complex metabolic reactions [37]. Maintaining high chlorophyll content in leaves is not necessary to improve the effective photosynthetic rate. Light intensity under low-light conditions is a limiting factor for leaf photosynthesis, and high chlorophyll content is conducive to light absorption; the photosynthetic rate under saturated light intensity is mainly affected by the catalytic ability of the Rubisco enzyme, rather than the limitation of electron transfer rate in light reactions [38]. To explore whether the increased chlorophyll content and abnormal leaf morphology of ltr1 affect photosynthetic efficiency, we measured the photosynthetic efficiency of NPB and ltr1 in the field. Compared to NPB, the intercellular CO 2 concentration of ltr1 was 4.91% higher, and the photosynthetic efficiency was 25.33% lower (Figure 2i,j). Although the photosynthetic pigment content of ltr1 increased, the photosynthetic efficiency did not. The reasons for the decrease of photosynthetic efficiency in ltr1 require further exploration.

Map-Based Cloning of LTR1
To explore the molecular mechanism of the phenotype in ltr1, an F 2 segregation population was developed by crossing ltr1 and the indica cultivar TN1. The segregation of wild type and mutant phenotype displayed a ratio of 3:1 (Table S1), indicating that the mutant phenotype was controlled by a single recessive gene. Using 21 F 2 mutant individuals, the LTR1 locus was first mapped to the region between RM6318 and RM1920 on the long arm of chromosome 2. The location was then narrowed down to a 13.5-kb genomic region between the markers N-12 and N-20 ( Figure 3a). In this region, only one putative opening reading frame (ORF) was found based on data from the Rice Genome Annotation Project (http://rice.plantbiology.msu.edu accessed on 24 March 2021) database.
DNA sequence analysis of the ORF in ltr1 and NPB revealed that a 2-bp deletion in exon 8 of LOC_Os02g40784, which resulted in a frameshift mutation and early termination of transcription ( Figure 3b). LOC_Os02g40784 includes ten exons and nine introns and encodes a polypeptide 619 amino acid in length. We therefore inferred that LOC_Os02g40784 was the gene controlling the mutant phenotype of ltr1.
the mutant phenotype was controlled by a single recessive gene. Using 21 F2 mutant individuals, the LTR1 locus was first mapped to the region between RM6318 and RM1920 on the long arm of chromosome 2. The location was then narrowed down to a 13.5-kb genomic region between the markers N-12 and N-20 ( Figure 3a). In this region, only one putative opening reading frame (ORF) was found based on data from the Rice Genome Annotation Project (http://rice.plantbiology.msu.edu) database. DNA sequence analysis of the ORF in ltr1 and NPB revealed that a 2-bp deletion in exon 8 of LOC_Os02g40784, which resulted in a frameshift mutation and early termination of transcription ( Figure  3b). LOC_Os02g40784 includes ten exons and nine introns and encodes a polypeptide 619 amino acid in length. We therefore inferred that LOC_Os02g40784 was the gene controlling the mutant phenotype of ltr1. To confirm that the phenotype of ltr1 was attributable to the detected mutation in LTR1, we constructed a complementation vector with a NPB genomic fragment containing the entire coding region of LTR1 and obtained complementary plants of LOC_Os02g40784 under ltr1 background. As expected, the complementary transgenic T 0 plants showed normal flat leaves: this indicated that the normal expression of LOC_Os02g40784 in ltr1 can complement the phenotype of the mutant (Figure 3c).

Overexpression and Targeted Deletion of LTR1
We next used CRISPR/Cas9 to generate mutant alleles of LTR1 alleles in a NPB background. We obtained three independent transgenic lines that all carried homozygous mutants, including 3-bp, 4-bp, and 5-bp deletions in exon 3, respectively (Figure 4a,e). These lines had comparable phenotypes to those of ltr1 with shrunken and distorted leaves, uneven distribution of bulliform cells on adaxial surface and sclerenchyma cells on abaxial surface, and disordered vascular bundles (Figure 4a-d). We also generated overexpression line of LTR1 in the NPB background, which exhibited longer leaves and higher relative expression level (Figure 5a-c). These results showed that LOC_Os02g40784 was LTR1 and that the mutation in LOC_Os02g40784 led to rumpled leaf phenotype in ltr1. Moreover, we found that compared with NPB, the grain yield per plant in overexpression of LTR1 increased by 38.59% (p < 0.05), but the grain yields per plant in ltr1 and LTR1-KO decreased by 82.13% and 76.31% (p < 0.05), respectively ( Figure S6), suggesting that overexpression of LTR1 enhanced yield in rice.
To confirm that the phenotype of ltr1 was attributable to the detected mutation in LTR1, we constructed a complementation vector with a NPB genomic fragment containing the entire coding region of LTR1 and obtained complementary plants of LOC_Os02g40784 under ltr1 background. As expected, the complementary transgenic T0 plants showed normal flat leaves: this indicated that the normal expression of LOC_Os02g40784 in ltr1 can complement the phenotype of the mutant (Figure 3c).

Overexpression and Targeted Deletion of LTR1
We next used CRISPR/Cas9 to generate mutant alleles of LTR1 alleles in a NPB background. We obtained three independent transgenic lines that all carried homozygous mutants, including 3-bp, 4-bp, and 5-bp deletions in exon 3, respectively (Figure 4a,e). These lines had comparable phenotypes to those of ltr1 with shrunken and distorted leaves, uneven distribution of bulliform cells on adaxial surface and sclerenchyma cells on abaxial surface, and disordered vascular bundles (Figure 4a-d). We also generated overexpression line of LTR1 in the NPB background, which exhibited longer leaves and higher relative expression level (Figure 5a-c). These results showed that LOC_Os02g40784 was LTR1 and that the mutation in LOC_Os02g40784 led to rumpled leaf phenotype in ltr1. Moreover, we found that compared with NPB, the grain yield per plant in overexpression of LTR1 increased by 38.59% (p < 0.05), but the grain yields per plant in ltr1 and LTR1-KO decreased by 82.13% and 76.31% (p < 0.05), respectively ( Figure S6), suggesting that overexpression of LTR1 enhanced yield in rice.  To examine the expression pattern of LTR1 in NPB, total RNA was extracted from roots, stem, leaf, sheath, and panicles. The qRT-PCR showed that LTR1 was constitutively expressed in all of the tested tissues, with a dramatic increase in leaves and panicles  To examine the expression pattern of LTR1 in NPB, total RNA was extracted from roots, stem, leaf, sheath, and panicles. The qRT-PCR showed that LTR1 was constitutively expressed in all of the tested tissues, with a dramatic increase in leaves and panicles ( Figure 5d). The results were consistent with those of β-glucuronidase (GUS) staining ( Figure 5e) and the decreased seed-setting rate of ltr1 (Figure 2f), showing the important regulatory role of LTR1 in leaf and panicle development.

LTR1 Participates in Water Transport and Ion Homeostasis
RNA-seq analysis showed that there were 6513 differentially expressed genes (DEGs) in NPB and ltr1, of which 3022 were up-regulated and 3480 were down-regulated ( Figure S2a and Table S6). There were 118 DEGs related to leaf development, comprising 36 upregulated and 82 down-regulated genes ( Figure S3a and Table S7). A Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis showed that these DEGs were mainly enriched in plant hormone signal transduction pathways, which indicated that LTR1 may regulate leaf development by participating in hormone signal transduction pathways ( Figure S3b

LTR1 Participates in Water Transport and Ion Homeostasis
RNA-seq analysis showed that there were 6513 differentially expressed genes (DEGs) in NPB and ltr1, of which 3022 were up-regulated and 3480 were down-regulated ( Figure S2a and Table S6). There were 118 DEGs related to leaf development, comprising 36 up-regulated and 82 down-regulated genes ( Figure S3a and Table S7). A Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis showed that these DEGs were mainly enriched in plant hormone signal transduction pathways, which indicated that LTR1 may regulate leaf development by participating in hormone signal transduction pathways ( Figure S3b A Gene Ontology (GO) term enrichment analysis was also conducted for DEGs between NPB and ltr1. The most highly enriched GO biological processes were in salt-stress response, stimulus response, and ABA response ( Figure S2b). The most highly enriched GO molecular functions were ATP binding and protein binding, and the most enriched cell components were plasma membrane and nucleus ( Figure S2c-d). These A Gene Ontology (GO) term enrichment analysis was also conducted for DEGs between NPB and ltr1. The most highly enriched GO biological processes were in salt-stress response, stimulus response, and ABA response ( Figure S2b). The most highly enriched GO molecular functions were ATP binding and protein binding, and the most enriched cell components were plasma membrane and nucleus ( Figure S2c,d). These results suggested that LTR1 was involved in the salt-stress response. It was previously reported that plant membrane transporters play key roles in resistance to biological and abiotic stresses; in particular, Na + /K + transporters increase resistance to salt stress [40]. We further found that there were 259 up-regulated and 178 down-regulated DEGs related to the salt-stress response (Figure 7a and Table S8). In ltr1, most of the genes encoding aquaporin or related to Na + /K + transporters were up-regulated, such as PLASMA MEMBRANE INTRINSIC PROTEIN genes OsPIP1;1, OsPIP1;2, OsPIP1;3, OsPIP2;1, OsPIP2;2, OsPIP2;4, OsPIP2;4; TONOPLAST INTRINSIC PROTEIN genes OsTIP1;1, OsTIP1;2; HIGH-AFFINITY K + TRANS-PORTERS genes OsHKT1;14, OsHKT2;3, and OsHKT1;5 (Figure 7b). These results suggested that LTR1 may affect salt tolerance by regulating water transport and ion homeostasis in plants through aquaporin and Na + /K + transporters. Given that many genes encoding aquaporins and ion transporter were differentially expressed in NPB and ltr1, we considered the possibility that LTR1 may regulate salt tolerance by affecting water transport and ion homeostasis. Therefore, we measured the Na + content in solution and in tissues of NPB and ltr1 under salt stress. After salt stress, Na + content in stems and leaves of ltr1 were significantly higher than those of NPB, which increased by 28.24% and 45.75%, respectively (p < 0.05). There was no significant difference in Na + content in the roots of NPB and ltr1 (p< 0.05) (Figure 7c). Furthermore, there was no significant difference in Na + content in the liquid media in which NPB and ltr1 plants were grown after treatment in hydroponic solution for 1 d (p < 0.05). However, after treatment for 3 or 6 d, Na + content was lower in the solution in which ltr1 plants were grown compared to NPB plants, decreased by 15.20% and 8.03%, respectively (p < 0.01) (Figure 7d). Under normal growth conditions (CK), the relative expression levels of OsPIP1;1, OsPIP1;2, OsPIP2;1, OsPIP2;2, OsTIP1;1, OsTIP1;2, OsHKT1;1, OsHKT1;5, and OsHKT2;3 were significantly higher in ltr1 than NPB leaves (Figure 7e), increased by 1.85, 2.36, 3.48, 3.46, 10.2, 2.57, 2.67, 2.52, 3.44 times, respectively (p < 0.01); which was consistent with the RNA-seq results. The genes encoding aquaporin and ion transporter in NPB and ltr1 plants both were strongly induced by salt stress. However, the induction of these genes was stronger in ltr1 than in leaves of NPB, leading to relative expression levels of OsPIP1;1, OsPIP1;2, OsPIP2;1, OsPIP2;2, OsTIP1;1, OsTIP1;2, OsHKT1;5, and OsHKT2;3 that were significantly higher in ltr1 than NPB leaves under salt stress (p < 0.01), especially the expression of OsPIP2;1 and OsHKT2;3 (Figure 7f). This was consistent with the finding that the Na + content in stems and leaves of ltr1 were significantly higher than those of NPB.

LTR1 Regulates Salt Tolerance in Rice
To further explore the function of LTR1 in the salt-stress response, we first screened a suitable salt concentration for treatment. NPB and ltr1 were cultured in soil treated with 0 mM NaCl (CK treatment), 50 mM NaCl, 100 mM NaCl, or 150 mM NaCl at the five-leaf stage. Two weeks later, the survival rates of NPB treated with 150 mM NaCl was higher than that of ltr1 plants (92.30% and 64.30%, respectively) (p < 0.05) ( Figure S4). We then grew NPB plants in solution, treated them with 150 mM NaCl, and measured the relative expression level of LTR1 at 0, 1, 3, 6, 12, and 24 h. The relative expression level of LTR1 increased overtime; the relative expression level of LTR1 increased by 6.95 times at 6 h and by 26.29 times at 24 h after treatment, indicating that LTR1 was significantly induced by salt stress (Figure 8c). After 7 d of salt stress in hydroponic solution, the survival rate of NPB reached 93.05%, which was significantly higher than that of ltr1 (43.52%) (p < 0.05) (Figure 8b). After 3 d of salt stress in hydroponic solution, H 2 O 2 and MDA in levels of NPB and ltr1 both accumulated, and the accumulation of MDA in the leaves of ltr1 was significantly higher than that of NPB (p < 0.05) (Figure 8d-f). These results suggested that, compared with NPB, the membrane lipid peroxidation and plasma membrane damage in ltr1 were more serious after salt stress, and that ltr1 was more sensitive to salt stress (Figure 8a-f). Studies have shown that when plants are subjected to stress, the enzymatic protection system is initiated rapidly, and the activities of peroxidase (POD), ascorbate peroxidase (APX), and other enzymes increase significantly, which enhances the capacity for reactive oxygen species (ROS) scavenging and reduces damage [41][42][43]. In this study, after salt stress, the catalase (CAT) activity in NPB and ltr1 decreased by 14.38% and 26.17%, respectively (p < 0.05). The decrease of CAT activity in ltr1 was more significant (Figure 8g). Furthermore, the activities of POD and APX in NPB and ltr1 both increased after stress, and the increases in POD and APX activities induced by stress in ltr1were weaker than that in NPB. After salt stress, the POD and APX activities of NPB increased by 32.31% and 81.62% compared with CK, while the POD and APX activities in ltr1 increased by 16.97% and 18.01% (p < 0.05) (Figure 8h,i). These were consistent with the expression change of antioxidant system in leaves of NPB and ltr1 (Figure 8j,k). Therefore, these results indicated that ltr1 had an inferior ability to adapt to salt stress.

LTR1 Encodes a Wax Synthesis Gene and Regulates Leaf Morphology
Cell structure is a key factor regulating leaf morphology. Many cloned genes regulated leaf morphology through affecting the normal development of vascular bundles, sclerenchyma cells, bulliform cells, epidermis, and cell walls [9]. However, few of these genes that affect leaf shape are involved in wax synthesis. In this study, we cloned a leaf shape gene, LEAF TIP RUMPLED1 (LTR1), which is an allele of the wax synthesis gene OsGL 1-4 [44]. LTR1 regulated leaf morphology, and loss function of LTR1 led to rumpled leaves with the abnormal development of bulliform cells, vascular bundles, and sclerenchyma cell. These indicated that LTR1 affected leaf morphology by regulating the development of bulliform cells, vascular bundles, and sclerenchyma cell. BR signal and auxin metabolism pathway played important roles in leaf morphogenesis [8,45]. OsBR6ox, which participates in brassinosteroid (BR) biosynthesis and signal transduction pathway, regulated normal development of organs and then induced abnormal vascular tissue and twisted leaves in its loss-of-function mutant [46]. OsARF16 [47] and OsARF17 [48] participate in the auxin response, affecting auxin polar transport and vascular tissue development. The RNA-dependent RNA polymerase OsRDR6 participates in formation of trans-acting small interfering RNA (ta-siRNA) [49], and ta-siRNA inhibits ARF3/ARF4 expression and thus inhibits maintenance of abaxial polarity [50]. OsAGO7, a ZIP/Ago7 homolog in Arabidopsis thaliana, is a critical member of the ta-siRNA-ARF3/ARF4-OsAGO7 complex and participates in regulation of leaf rolling [51]. In this study, OsBR6ox, OsARF16, OsARF17 and OsRDR6 were found to be differentially expressed in NPB and ltr1. We therefore speculated that LTR1 may affect leaf morphology by participating in plant hormone signal transduction pathway, while the detailed regulatory network involved requires further study.

LTR1 Has Multiple Effects on Plant Growth and Development
There are 11 Glossy1 (GL1) homologous genes in rice, OsGL1-1 through OsGL1-11, which vary expression levels between rice tissues and organs. Most are induced by abiotic stress and play key roles in wax synthesis and stress tolerance [44]. It was reported that OsGL1-1, OsGL1-2, OsGL1-3, and OsGL1-6 affect the leaf water loss rate by controlling the wax content in the leaf epidermis, thereby controlling drought resistance in rice [31,44,52,53]. In the present study, we found that LTR1, an allele of OsGL1-4, was also involved in the regulation of salt tolerance with LTR1 strongly induced by salt stress. The ltr1 plants showed high sensitivity to salt stress compared to the wild-type, with more serious membrane lipid peroxidation and plasma membrane damage. Moreover, in rice, many humidity-sensitive genic male sterile lines (HGMS) were obtained by identifying wax synthesis genes involved in regulating pollen development and affecting panicle fertility. Previous studies have shown that most wax synthesis genes, such as DROUGHT HYPERSENSITIVE (DPS1) [32], SUBTILISIN-LIKE SERINE PROTEASE 1 (SUBSrP1) [54], HMS1-INTERACTING PROTEIN (HMS1I) [55], HUMIDITY-SENSITIVE GENIC MALE STERILITY 1 (HMS1) [56], and OsGL1-5 [44] were involved in the regulation of panicle fertility. Loss functions of these genes resulted in abnormal pollen development and a decrease in the seed setting rate at low humidity but a normal seed setting rate at high humidity. Based on this mechanism, the corresponding mutants can be used as HGMSs. It has also been reported that OsGL1-4 controls male sterility in rice by affecting pollen adhesion and water cooperation under ambient humidity [57]. We here found that loss function of LTR1 resulted in a severe decrease in the seed setting rate and grain yield per plant, and significant changes in the number of branches and effective panicles in ltr1. What's more, overexpression of LTR1 enhances yield in rice. These results indicated that LTR1 had pleiotropic functions in rice growth and development.

LTR1 Regulated Salt Tolerance by Altering Plant Water Status and Ion Homeostasis
Plant aquaporins play very important roles in water transport of transmembrane and form a large protein family [58]. Great progress has been made in functional studies of plasma membrane intrinsic proteins (PIPs) and tonoplast intrinsic proteins (TIPs), which have shown that their main physiological function is to promote transmembrane transport of osmotic water [58]. The expression regulation of PIPs varies with differing experimental conditions [59]. OsPIP1;1 showed low water channel activity in Xenopus oocytes, but the permeability of OsPIP1;1 improved significantly when it was co-expressed with OsPIP2.1 [60]. In the present study, the relative expression level of OsPIP2;1 was much higher than that of OsPIP1;1, OsPIP1;2, and OsPIP2;2). This indicated that the upregulation of OsPIP2;1 resulted in enhanced leaf permeability and poor water retention in ltr1. Class I HKT transporters play an important role in removing sodium ions from the xylem [61,62]. Because the accumulation of K + in plant cells homeostasis the toxicity of Na + accumulation, stable acquisition and distribution of K + are required during salt-stress conditions [63]. The OsHKT transporter is involved in Na + transport in rice, and OsHKT1 specifically mediates Na + uptake by rice roots under conditions of K + deficiency [64]. OsHKT1;5 controls the transport of K + and Na + from roots to shoots. Under salt stress, OsHKT1;5 refluxes of excessive Na + from shoots to roots by unloading it from the xylem, thereby reducing Na + toxicity and enhancing salt tolerance [61]. However, we here found that high expression of OsHKT1;5 under high salt conditions did not reduce the accumulation of Na + in ltr1 leaves. Thus, the excessive accumulation of Na + in ltr1 under salt stress may be regulated by other factors. Under salt stress, the relative expression of HKT2;3 was significantly higher than the expression of other genes encoding ion transporters. Meanwhile, overexpression of the aquaporin gene OsPIP2;1 led to enhanced water permeability and poor water retention in ltr1. More Na + was absorbed by ltr1 than NPB roots and transported to aboveground parts; thus, the Na + content was significantly higher in stems and leaves of ltr1 than NPB. Furthermore, there were more white crystals on the stems of NPB than that of ltr1 ( Figure S5). These results suggested that over-accumulation of Na + in ltr1 could not be reversed in a timely fashion, resulting in high sensitivity of ltr1 to salt stress. Therefore, we speculated that LTR1 affected the water status and ion homeostasis of plants by regulating the expression of genes encoding aquaporins and ion transporters, which ultimately regulated salt tolerance in plants.

Prospects
Wax, cuticle, and polysaccharide form the cuticle of epidermis, which is a selfprotective barrier against biotic and abiotic stresses in plants [29,65,66]. Wax affects canopy temperature and water transport in plants, which further affect plants adaptation to harmful environmental factors such as heat/drought/salt stress and pest/pathogen damage [29,32]. Here, we found that the wax synthesis gene LTR1 regulates leaf morphology by affecting the normal development of bulliform cells, vascular bundles, and sclerenchyma cells. Moreover, overexpression of LTR1 enhanced yield in rice and LTR1 positively regulates salt stress by affecting water and ion homeostasis in plants. However, the regulatory and response mechanism by which LTR1 affected leaf morphogenesis, water retention, and ion transport between the root and shoot requires further analysis. The differences in ion transport (ion flow rate, ion transport efficiency) and horizontal balance ability between NPB and ltr1, together with their regulatory mechanisms need to be further analyzed. How wax content affects cell structure, tissue moisture, and ion balance need further exploration. Identifying proteins that directly interact with LTR1 and analyzing the molecular mechanism of their interaction in regulating leaf shape and salt tolerance will further supplement the known genetic regulation network that governs leaf shape and salt tolerance, providing a theoretical foundation for breeding high-yield rice varieties with high salt tolerance. In addition, identification and application of favorable alleles of LTR1, which confers resistances without negative effects on yield, can potentially be used to breed high-yield and high-resistance rice varieties through the combination of multi-omics and bioinformatics. Therefore, according to the insights uncovered in this study, LTR1 can be considered as a potentially highly valuable gene resource for the improvement of leaf morphology and stress resistance in rice breeding. Manipulating genes associated with leaf morphology and stress resistance individually or in combination makes it possible in the "precision breeding" to breed rice varieties with ideal plant architecture and high resistances without yield penalties. Thus, our results illustrate innovative approaches for developing potentially high stress resistant crop varieties with ideal plant architecture and carry significant implications for breeding application of high yield and stress-resistance-related genetic resources.

Plant Materials and Growth Conditions
In this study, the ltr1 mutant was isolated from a population of the Oryza sativa ssp. japonica variety Nipponbare (NPB) mutagenized with a 1% ethyl methanesulfonate (EMS) solution using a forward genetic screen for altered leaf shape.

Phenotypicl Characterization and Histological Analysis
To investigate whether the LTR1 mutation affected rice yield, agronomic traits such as panicle length, effective panicle number, numbers of branches, grain numbers per panicle, seed setting rate, grain yield per plant, and 1000-grain weight were measured for each of 5 or 6 biological replicates at the mature stage. The panicle length, number of branches, and grain numbers per panicle were obtained from measurements of the main panicle.
For frozen cross-section assays, the leaves were immersed in frozen embedding agent (Tissue-Tek ® O.C.T. Compound, Sakura, Tokyo, Japan) for 2-3 h at -20 • C. Sections (15 µm) were cut with a freezing microtome (Leica CM1950, Wetzlar, Germany) and placed on microscope slides. Slices were observed and photographed using a microscope (Leica DM4 B). The areas of bulliform cells were calculated using Image J software.

Measurements of Chlorophyll Content and Photosynthetic Parameters
Chlorophyll a, Chlorophyll b, and carotenoid (Car) content were measured in three biological replicates using the methods described by Sartory and Grobbelaar [67].
SPAD values were determined for ten biological replicates using a SPAD-502 PLUS. Chlorophyll fluorescence was measured for ten biological replicates with a FluorPen FP100. The QY (Fv/Fm) was determined after a 20 min dark adaptation period.
The net photosynthesis rate, stomatal conductance, and transpiration rate of NPB and ltr1 plants were evaluated for eight biological replicates with a Li-COR 6400 portable system. All measurements were conducted under the following conditions: photosynthetic photon flux density of 1200 µmol·m −2 ·s −1 , ambient CO 2 (400 µmol·mol −1 ), 6 cm 2 of leaf area, 500 µmol·s −1 flow speed, and ambient temperature.

Map-Based Cloning and Complementation Assay
To fine-map the mutated gene, an F 2 population was constructed from a cross between ltr1 and a wild-type indica variety, TN1, with flat leaves. Plants from this population that exhibited rumpled leaves were selected for gene mapping. The locus was first mapped to an interval between the two markers RM6318 and RM1920 (Table S2) on the long arm of chromosome 2, then was further narrowed down to a 13.5-kb DNA region. There was only one open reading frame (ORF) in this region. Genomic DNA fragments in this region were amplified using primers listed in Table S3 from NPB and ltr1.
An 8628-bp genomic DNA fragment containing the coding region of LOC_Os02g40784, plus 2060-bp upstream, 5450-bp of the coding region and 1118-bp downstream regions, was amplified from NPB (primers for this process are shown in Table S4) and then was cloned into the binary vector pCAMBIA1300 by homologous recombination. The resulting construct pCAMBIA1300-LTR1 was transformed into ltr1 calli to obtain complementary transgenic plants.

Gene Editing and Overexpression
For generation of knockout plants using CRISPR/Cas9 technology, gene-specific guide sequences (primers are listed in Table S4) targeting LTR1 were designed to create single guide RNAs (sgRNAs), after which the sgRNA-Cas9 sequences were cloned into pYLCRISPR/Cas9-MH [68].
Full-length cDNA of LTR1 amplified (primers are listed in Table S4) from NPB was cloned into the Gateway entry vector pDONR ZEO (Invitrogen, Carlsbad, CA, USA), then recombined into the pUbi::attR-GFP-3×FLAG vector using the Gateway cloning system (Invitrogen). The resulting construct was transformed into NPB calli to obtain overexpression lines of LTR1.

Histological GUS Assay
The promoter of LTR1 (2094-bp upstream of the start codon) was amplified from NPB genomic DNA (primers are listed in Table S4) and inserted into the EcoRI and NcoI sites of the binary vector pCAMBIA1305.1. This resulted in a fusion of the promoter and the GUS reporter gene (pLTR1::GUS). The recombinant vector was then introduced into NPB calli to obtain transgenic plants.

RNA-seq and Data Analysis
Plants were harvested for total RNA extraction at the booting stage. Three biological replicates were used for RNA-seq analysis. The RNA-seq libraries were constructed and sequenced using an Illumine HiSeq. Each sample obtained approximately 20,000,000 clean reads, which were mapped to NPB reference genome based on the genome information by HISAT2 (http://ccb.jhu.edu/software/hisat2/index.shtml accessed on 10 July 2022). Differential expression analysis for NPB and ltr1 was performed with DESeq2 using thresholds of FDR < 0.01 and |log 2 (fold change)| ≥ 2). A GO enrichment analysis was implemented with the GOseq R packages. The KEGG pathway analysis of DEGs was conducted using the KEGG database (http://www.genome.jp/kegg/ accessed on 24 March 2021).

Determination of Stress-Related Physiological Index
For 3,3 -diaminobenzidine (DAB) staining, 0.1 g DAB was fully dissolved in ddH 2 O by adjusting pH to 5.8. The samples were immersed into a tube containing 1 mg/mL DAB solution overnight at 28 • C under dark conditions. Stained samples were cleared of chlorophyll by dehydration with 80% ethanol, then scanned using a Microtek Scan Maker i800 plus. The contents of hydrogen peroxide (H 2 O 2 ) and malondialdehyde (MDA), activities of catalase (CAT), peroxidase (POD), and ascorbate peroxidase (APX) were measured using appropriate kits from Geruisi (http://www.geruisi-bio.com/ accessed on 18 May 2021) following the manufacturer's instructions with four biological replicates per sample.

Measurement of Na + Content
For each sample, a total of 0.05 g of dry tissues power were weighed and immersed in 4 mL concentrated nitric acid with 2 mL 30% H 2 O 2 overnight, and then were decocted with temperature gradient (60 • C for 1 h, 120 • C for 1 h, 160 • C for 1 h, 190 • C until the solution was clarified) using a graphite digestion instrument (DigiBlock ED54, Beijing, China). The content of extracted Na + was measured by inductively coupled plasma-mass spectrometry (ICP-MS) (iCAP RQ, Thermo Fisher Scientific, 168 Third Avenue, Waltham, MA, USA) after acid catching, constant volume and filtration with three biological replicates per sample.

RNA Extraction and Quantitative Real-Time PCR
Total RNA was extracted using the Total RNA Miniprep kit (Axygen, Hangzhou, China) following the manufacturer's instructions. First-strand cDNA was synthesized using the ReverTra Ace qPCR-RT kit (Toyobo, Osaka, Japan) as instructed by the manufacturer, using 2 µg of total RNA for each reaction. qRT-PCR analyses were performed using SYBR Premix Ex Taq (Takara, Kusatsu, Japan) and gene-specific primers on a CFX96TM real-time system. Three or four biological replicates were performed for all experiments. The primers used are listed in Table S5 in the Supplementary Materials.

Quantification and Statistical Analysis
Quantification analyses on all the measurements were conducted in GraphPad Prism 8. Significant differences were determined with Student's t-test and Duncan's new multiple range tests.