*2.7. 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, H2O<sup>2</sup> 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 [41–43]. 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 *ltr1*were 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.

**Figure 7.** *LTR1* regulates salt-stress response by regulating genes encoding aquaporins and ion transporters. (**a**) Volcano plot of DEGs related to salt response between NPB and *ltr1*. (**b**) Heat map of significantly up-regulated DEGs encoding aquaporin and ion transporters between NPB and *ltr1*. (**c**) Na<sup>+</sup> content in different tissues of NPB and *ltr1*. (**d**) Na<sup>+</sup> content in solutions where NPB and *ltr1* were cultured after treatment for 1, 3, or 6 d. (**e**) Relative expression levels of *LTR1* and genes encoding aquaporin and ion transporters under normal condition (CK). (**f**) The relative expression levels of *LTR1* and genes encoding aquaporin and ion transporters under 150 mM NaCl (Salt). Data **Figure 7.** *LTR1* regulates salt-stress response by regulating genes encoding aquaporins and ion transporters. (**a**) Volcano plot of DEGs related to salt response between NPB and *ltr1*. (**b**) Heat map of significantly up-regulated DEGs encoding aquaporin and ion transporters between NPB and *ltr1*. (**c**) Na<sup>+</sup> content in different tissues of NPB and *ltr1*. (**d**) Na<sup>+</sup> content in solutions where NPB and *ltr1* were cultured after treatment for 1, 3, or 6 d. (**e**) Relative expression levels of *LTR1* and genes encoding aquaporin and ion transporters under normal condition (CK). (**f**) The relative expression levels of *LTR1* and genes encoding aquaporin and ion transporters under 150 mM NaCl (Salt). 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* < 0.05; \*\* in the figure represents significant difference at *p* < 0.01 and ns in the figure represents there is no significant difference at *p* < 0.05.

**Figure 8.** The response of *LTR1* to salt stress in NPB and *ltr1*. (**a**) Photos of NPB and *ltr1* under CK and Salt treatment, bar = 10.5 cm. (**b**) The survival rate of NPB and *ltr1* after treatment for 7 d. (**c**) The relative expression level of *LTR1* after treatment for 0, 1, 3, 6, 12, 24 h. (**d**) DAB staining in leaves of NPB and *ltr1* under CK and salt treatment. (**e**,**f**) MDA and H2O<sup>2</sup> content in leaves of NPB and *ltr1* under CK and Salt treatment. (**g**–**i**) CAT, POD, and APX activity in leaves of NPB and *ltr1* under CK and salt treatment. (**j**,**k**) The relative expression level of genes related to antioxidant system in leaves of NPB and *ltr1* under CK and Salt treatment, *n* = 4. 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* < 0.05 and ns in the figure represents there is no significant difference at *p* < 0.05. Different lowercase letters indicate significant differences based on the Duncan's new multiple range test (*p* < 0.05).

## **3. Discussion**

### *3.1. 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.

#### *3.2. 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.

#### *3.3. 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<sup>+</sup> in plant cells homeostasis the toxicity of Na<sup>+</sup> accumulation, stable acquisition and distribution of K<sup>+</sup> are required during salt-stress conditions [63]. The OsHKT transporter is involved in Na<sup>+</sup> transport in rice, and OsHKT1 specifically mediates Na<sup>+</sup> uptake by rice roots under conditions of K<sup>+</sup> deficiency [64]. *OsHKT1;5* controls the transport of K<sup>+</sup> and Na<sup>+</sup> from roots to shoots. Under salt stress, *OsHKT1;5* refluxes of excessive Na<sup>+</sup> from shoots to roots by unloading it from the xylem, thereby reducing Na<sup>+</sup> 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<sup>+</sup> in *ltr1* leaves. Thus, the excessive accumulation of Na<sup>+</sup> 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<sup>+</sup> was absorbed by *ltr1* than NPB roots and transported to aboveground parts; thus, the Na<sup>+</sup> 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<sup>+</sup> 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.

### *3.4. 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.

#### **4. Materials and Methods**

### *4.1. 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. Rice plants were grown under natural environmental conditions in an experimental field at the China National Rice Research Institute in Fuyang District (Zhejiang province, China) and Lingshui (Hainan province, China).

Seedlings used in salt treatments were cultured in soil and hydroponic solution (1.25 mM NH4NO3, 0.3 mM KH2PO4, 0.35 mM K2SO4, 1 mM CaCl2, 1 mM MgSO4, 0.5 mM NaSiO3·9H2O, 20 µM Fe-EDTA, 9 µM MnCl2·4H2O, 0.39 µM (NH4)6Mo7O24·4H2O, 20 µM

H3BO3, 0.77 µM ZnSO4·7H2O, 0.32 µM CuSO4·5H2O) in an artificial incubator with a 12 h/12 h light/dark at 70–80% humidity and a 25–30 ◦C/28 ◦C day/night temperature (MLR-352H-PC, Panasonic, Osaka, Japan). For salt-stress treatments, plants were cultivated in hydroponic media containing 0 mM NaCl (CK), 50 mM NaCl, 100 mM NaCl, or 150 mM NaCl (Salt), respectively.

#### *4.2. 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.

#### *4.3. 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 <sup>µ</sup>mol·m−<sup>2</sup> ·s −1 , ambient CO<sup>2</sup> (400 <sup>µ</sup>mol·mol−<sup>1</sup> ), 6 cm<sup>2</sup> of leaf area, 500 µmol·s <sup>−</sup><sup>1</sup> flow speed, and ambient temperature.

#### *4.4. Map-Based Cloning and Complementation Assay*

To fine-map the mutated gene, an F<sup>2</sup> 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.

#### *4.5. 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*.

#### *4.6. 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 *Eco*RI and *Nco*I 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.

For GUS staining, different tissues of transgenic plants were incubated in X-Gluc buffer (0.1 mol·L <sup>−</sup><sup>1</sup> <sup>K</sup>2HPO<sup>4</sup> (pH 7.0), 0.1 mol·<sup>L</sup> <sup>−</sup><sup>1</sup> KH2PO<sup>4</sup> (pH 7.0), 5 mmol·<sup>L</sup> <sup>−</sup><sup>1</sup> K3Fe (CN)6, 5 mmol·L <sup>−</sup><sup>1</sup> <sup>K</sup>4Fe (CN)6·3H2O, 0.1% Triton X-100, 20% methanol, and 1 mg·mL−<sup>1</sup> X-Gluc) at 37 ◦C for 2 h [69]. Stained samples were cleared of chlorophyll by dehydration with ethanol, then scanned using a Microtek Scan Maker i800 plus.

#### *4.7. 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<sup>2</sup> (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).

#### *4.8. Determination of Stress-Related Physiological Index*

For 3,30 -diaminobenzidine (DAB) staining, 0.1 g DAB was fully dissolved in ddH2O 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 (H2O2) 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.
