LMM24 Encodes Receptor-Like Cytoplasmic Kinase 109, Which Regulates Cell Death and Defense Responses in Rice

Lesion mimic mutants are excellent models for research on molecular mechanisms of cell death and defense responses in rice. We identified a new rice lesion mimic mutant lmm24 from a mutant pool of indica rice cultivar “ZhongHui8015”. The LMM24 gene was identified by MutMap, and LMM24 was confirmed as a receptor-like cytoplasmic kinase 109 by amino acid sequence analysis. The lmm24 mutant displayed dark brown lesions in leaves and growth retardation that were not observed in wild-type ZH8015. The results of histochemical staining and TUNEL assays showed enhanced ROS accumulation and cell death in lmm24. Chloroplast degradation was observed in lmm24 leaves, with decreased expression of photosynthesis-related genes and increased expression of the senescence-induced STAYGREEN (SGR) gene and other senescence-associated genes. Furthermore, lmm24 exhibited enhanced resistance to rice blast fungus Magnaporthe oryzae (M. oryzae) and up-regulation of defense response genes. Our data demonstrate that LMM24 regulates cell death and defense responses in rice.


Introduction
Lesion mimic mutants (LMMs) display spontaneous lesions in the absence of pathogen attack, environmental stress, or mechanical damage [1], exhibiting a similar phenotype to the pathogen infection-induced hypersensitive response (HR) mediated through programmed cell death (PCD) [2]. Programmed cell death plays an important role in the innate immunity of plants by initiating the cell death process to defend against pathogen infection [3]. In fact, most LMMs show increased resistance to pathogens: The spl30 mutant shows enhanced resistance to the bacterial pathogen Xanthomonas oryzae pv oryzae (Xoo); dj-lm shows enhanced resistance to rice blast (M. oryzae) [4]; and oscul3a shows enhanced resistance to both M. oryzae and Xoo [5,6]. These findings suggest that LMM genes are involved in regulating the plant defense response. Cloning and characterization of LMM genes will likely provide new insights into complex defense pathways in plants.
Lesion mimic mutants have been reported in various plants, such as Arabidopsis (Arabidopsis thaliana) [7], maize (Zea mays L.) [8], rice (Oryza sativa L.) [9], wheat (Triticum aestivum L.) [10], and barley (Hordeum vulgare L.) [11]. In recent years, more and more LMM genes with diverse pathogenesis of lesions have been isolated and characterized in rice. For example, NLS1 encodes an ancient class of (dps)) in the field in summer. Individuals of lmm24 grown in the greenhouse showed lesions at 20 dps (28 °C, 8 h dark, 16 h light), which then spread from the leaf tip to the entire leaf. However, prior to 25 dps, the leaves of the mutant had no lesions (similar to ZH8015 leaves). Moreover, the newly emerging leaves showed no lesions but gradually developed lesions as the plant developed ( Figure  1B,C). There were fewer lesions on the flag leaf at the maturation period, which may be related to the temperature and level of illumination of the environment ( Figure 1E).
Numerous studies on LMMs have shown that light is an important factor affecting the occurrence of lesions. In order to determine whether lesion formation in lmm24 was affected by light, we conducted an antiglare assay on both ZH8015 and lmm24. The same size of aluminum foil was used to cover newly emerging leaves of the mutant and ZH8015 to prevent light exposure ( Figure  1D). In lmm24, areas without exposure to light did not have lesions compared to areas exposed to light. There were no differences between the two parts in ZH8015. Thus, we determined that lmm24 develops lesions in a light-dependent manner.
In addition, compared with ZH8015, plant height, tiller number, grain size, 1000 grain weight, and seed setting rate of lmm24 were significantly reduced (Figures 1F,G and S1).  Numerous studies on LMMs have shown that light is an important factor affecting the occurrence of lesions. In order to determine whether lesion formation in lmm24 was affected by light, we conducted an antiglare assay on both ZH8015 and lmm24. The same size of aluminum foil was used to cover newly emerging leaves of the mutant and ZH8015 to prevent light exposure ( Figure 1D). In lmm24, areas without exposure to light did not have lesions compared to areas exposed to light. There were no differences between the two parts in ZH8015. Thus, we determined that lmm24 develops lesions in a light-dependent manner.
In addition, compared with ZH8015, plant height, tiller number, grain size, 1000 grain weight, and seed setting rate of lmm24 were significantly reduced ( Figure 1F,G and Figure S1).

Abnormal ROS Accumulation and Cell Death in lmm24
We speculated that the lesion phenotype of LMMs is the result of ROS accumulation leading to cell death. Thus, we investigated ROS accumulation and cell death in lmm24 leaves. Evans blue solution can enter cells undergoing death but cannot enter living cells, so Evans blue staining is a histochemical indicator of cell death. The reagent 3,3 -diaminobenzidine (DAB) was used to indicate the accumulation of hydrogen peroxide (H 2 O 2 ). Histochemical staining results showed that lmm24 had many Evans blue spots after staining, whereas ZH8015 did not (Figure 2A), indicating mass cell death in lmm24. Comparing DAB staining between the mutant and ZH8015, we found that lmm24 had much stronger H 2 O 2 accumulation than ZH8015 ( Figure 2B). Terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) assay is another method to detect cell death. When the cell dies, the DNA breaks and the exposed 3'-OH can be added to fluorescein by the TUNEL reagent, which can be detected by fluorescence microscopy. Our TUNEL assay carried out to detect DNA fragmentation in lmm24 and ZH8015 indicated stronger TUNEL signals were detected in the leaves of lmm24 ( Figure 2C), which also suggests that there was much more cell death in lmm24.

Abnormal ROS Accumulation and Cell Death in lmm24
We speculated that the lesion phenotype of LMMs is the result of ROS accumulation leading to cell death. Thus, we investigated ROS accumulation and cell death in lmm24 leaves. Evans blue solution can enter cells undergoing death but cannot enter living cells, so Evans blue staining is a histochemical indicator of cell death. The reagent 3,3′-diaminobenzidine (DAB) was used to indicate the accumulation of hydrogen peroxide (H2O2). Histochemical staining results showed that lmm24 had many Evans blue spots after staining, whereas ZH8015 did not (Figure 2A), indicating mass cell death in lmm24. Comparing DAB staining between the mutant and ZH8015, we found that lmm24 had much stronger H2O2 accumulation than ZH8015 ( Figure 2B). Terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) assay is another method to detect cell death. When the cell dies, the DNA breaks and the exposed 3'-OH can be added to fluorescein by the TUNEL reagent, which can be detected by fluorescence microscopy. Our TUNEL assay carried out to detect DNA fragmentation in lmm24 and ZH8015 indicated stronger TUNEL signals were detected in the leaves of lmm24 ( Figure 2C), which also suggests that there was much more cell death in lmm24.

Early Leaf Senescence in lmm24
A common feature of most LMMs is that leaf senescence is a consequence of uncontrolled PCD. The degradation of chloroplasts is a critical indicator of leaf senescence, so we compared the ultrastructure of chloroplasts in lmm24 and ZH8015 using transmission electron microscopy (TEM). Chloroplasts were well-developed with rich lamellae and a small number of osmiophilic bodies in ZH8015 ( Figure 3A-C), whereas chloroplast membranes were broken and the thylakoid lamellae had loosely arranged mesophyll cells surrounding the lesion areas in lmm24 ( Figure 3D-F). Expression levels of photosynthesis-related genes (psaB, psbA, psbB, psbC, cab2R, rpoA, CHLI, CHLD) were significantly down-regulated in lmm24 mutant leaves ( Figure 3G), while expression levels of the senescence-induced STAYGREEN (SGR) gene and senescence-associated genes (Osl2, Osl30, Osl43, Osl85) were significantly up-regulated ( Figure 3H). At the same time, Ubiquitin was used as a reference gene to analyze the relative expression levels of the above genes, and the results were consistent with the results of β-actin1 as a reference gene ( Figure S2). Overall, these results suggest that early leaf senescence occurs in lmm24.

Early Leaf Senescence in lmm24
A common feature of most LMMs is that leaf senescence is a consequence of uncontrolled PCD. The degradation of chloroplasts is a critical indicator of leaf senescence, so we compared the ultrastructure of chloroplasts in lmm24 and ZH8015 using transmission electron microscopy (TEM). Chloroplasts were well-developed with rich lamellae and a small number of osmiophilic bodies in ZH8015 ( Figure 3A-C), whereas chloroplast membranes were broken and the thylakoid lamellae had loosely arranged mesophyll cells surrounding the lesion areas in lmm24 ( Figure 3D-F). Expression levels of photosynthesis-related genes (psaB, psbA, psbB, psbC, cab2R, rpoA, CHLI, CHLD) were significantly down-regulated in lmm24 mutant leaves ( Figure 3G), while expression levels of the senescence-induced STAYGREEN (SGR) gene and senescence-associated genes (Osl2, Osl30, Osl43, Osl85) were significantly up-regulated ( Figure 3H). At the same time, Ubiquitin was used as a reference gene to analyze the relative expression levels of the above genes, and the results were consistent with the results of β-actin1 as a reference gene ( Figure S2). Overall, these results suggest that early leaf senescence occurs in lmm24.

lmm24 Shows Enhanced Resistance to M. oryzae
Previous studies indicated that some LMMs show enhanced resistance to fungal and bacterial pathogens. To determine whether lmm24 also confers higher resistance to pathogens, we employed the leaf spraying (seedling stage) and punch inoculation (tillering stage) methods to infect ZH8015 and lmm24 plants with the virulent M. oryzae isolate 12-144-1-1 ( Figure 4A,B). Seedlings of both the mutant and ZH8015 showed susceptibility, but the morbidity rate of ZH8015 was more serious than that of the mutant. ZH8015 was obviously susceptible at the tillering stage while the mutant was minimally susceptible to disease. The statistical results of lesion area in the two growth stages showed that ZH8015 had significantly higher areas with lesions than the mutant ( Figure 4C,D), indicating that lmm24 enhanced resistance to M. oryzae. To determine the mechanism of enhanced resistance of lmm24, we investigated the transcription level of PR genes (PR1a, PBZ1, PR1b, PAL1, AOS2, WRKY45) in ZH8015 and lmm24 ( Figure 4E,F); these genes were all significantly upregulated in lmm24. The same experiment was performed using Ubiquitin as a reference gene. Expression levels of PR genes were also significantly greater in lmm24 than in ZH8015 ( Figure S3). Together, these results suggested that the lmm24 mutant gene triggered a defense response to M. oryzae, which led to enhanced disease resistance associated with lesion formation in lmm24.

Genetic Analysis and Use of the MutMap Method to Clone the LMM24 Gene
To isolate the potential causal gene of the observed mutant phenotype, we crossed lmm24 with ZH8015. F 1 plants were self-pollinated to obtain F 2 progeny. All F 1 plants exhibited a ZH8015 phenotype while the F 2 progeny exhibited a ZH8015/mutant ratio of 3:1 (Table S1), suggesting that the phenotype of lmm24 was controlled by a single recessive nuclear gene.
The MutMap method was used to map the lmm24 gene. From the F 2 population, 30 individual plants with mutant phenotypes were selected to extract DNA which were pooled (pool-M) for genome-wide resequencing. We obtained 22,155,026,400 clean base pairs from pool-M and 6,754,683,300 clean base pairs from ZH8015 pool (pool-WT). Compared with the reference genome, the mapping rate of pool-M was 97.78% and the average depth was 50×; the mapping rate of pool-WT was 97.69% and average depth was 16× (Table S3 and Table S4). These results indicate that the quality and depth of sequencing sufficient for the subsequent SNP analysis.
We used the UnifiedGenotyper module of the GATK3.3 (Broad Institute of MIT and Harvard, Cambridge, MA, USA) software to detect multiple sample SNPs, filtered using VariantFiltration, and annotated SNP results with ANNOVAR (Columbia University, New York, NY, USA). According to the method described by Akira Abe et al. [29], we calculated the SNP index of each SNP and plotted the SNP distributions of 12 chromosomes in rice ( Figure S4). Regions with a SNP index of 1 are considered candidate regions for the lmm24 mutant phenotype. Fortunately, only one region on chromosome 3 was selected as a candidate, corresponding to the rice gene BGIOSGA012679 (LOC_Os03g24930) ( Figure 5A). We sequenced the cDNA of LOC_Os03g24930 and found two nucleotide mutations (C 1009 -to-T, C 1055 -to-T) and a 54 bp insertion; these two loci were in the fourth exon of this gene leading to amino acid changes ( Figure 5B). In order to verify the 54 bp insertion, we extracted the DNA of 15 individual plants with mutant phenotypes in the F 2 group and carried out agarose gel electrophoresis using a molecular marker at the insertion site (4th Exons F/R, Table S2). The results showed that the 54 bp insertion was present in all 15 individual plants ( Figure 5C). Thus, LOC_Os03g24930 is a candidate gene for LMM24.
nucleotide mutations (C 1009 -to-T, C 1055 -to-T) and a 54 bp insertion; these two loci were in the fourth exon of this gene leading to amino acid changes ( Figure 5B). In order to verify the 54 bp insertion, we extracted the DNA of 15 individual plants with mutant phenotypes in the F2 group and carried out agarose gel electrophoresis using a molecular marker at the insertion site (4 th Exons F/R, TableS2). The results showed that the 54 bp insertion was present in all 15 individual plants ( Figure 5C). Thus, LOC_Os03g24930 is a candidate gene for LMM24. The fragment corresponding to ZH8015 is the size of the PCR product without insertion mutation. The fragment corresponding to lmm24 is the size of the PCR product with the insertion mutation.

Functional Complementation with LOC_Os03g24930 in the lmm24
To verify that LOC_Os03g24930 is the LMM24 gene, the vector COM-lmm24 containing a 5192 bp genomic fragment of ZH8015, consisting of the 1842 bp upstream promoter and 1166 bp downstream terminator of LOC_Os03g24930, was constructed. This vector was then introduced into lmm24 by Agrobacterium tumefaciens-mediated transformation. Of 32 regenerated T 0 plants, 27 were positive transformants, and all of them were similar in phenotype to ZH8015 and thus the mutant phenotype of lmm24 was restored. There were no lesions on the leaves ( Figure 6A

LMM24 Encodes Receptor-Like Cytoplasmic Kinase 109 in Rice
As annotated on RGAP (http://rice.plantbiology.msu.edu), LOC_Os03g24930 (LMM24) is composed of four exons and three introns, a genomic DNA (gDNA) sequence of 2185 bp, a coding sequence of 1344 nucleotides, and encodes a putative 448 amino acid protein with a tyrosine protein kinase domain. Previous research showed that LMM24 encodes receptor-like cytoplasmic kinase 109 (OsRLCK109) [18]. RLCKs in plants belong to the super family of receptor-like kinases (RLKs), which lack extracellular ligand-binding domains. The predicted protein scheme for OsRLCK109 shows a

LMM24 Encodes Receptor-Like Cytoplasmic Kinase 109 in Rice
As annotated on RGAP (http://rice.plantbiology.msu.edu), LOC_Os03g24930 (LMM24) is composed of four exons and three introns, a genomic DNA (gDNA) sequence of 2185 bp, a coding sequence of 1344 nucleotides, and encodes a putative 448 amino acid protein with a tyrosine protein kinase domain. Previous research showed that LMM24 encodes receptor-like cytoplasmic kinase 109 (OsRLCK109) [18]. RLCKs in plants belong to the super family of receptor-like kinases (RLKs), which lack extracellular ligand-binding domains. The predicted protein scheme for OsRLCK109 shows a kinase domain ( Figure 7A); 108-393 amino acids encode the kinase domain. Multiple series of alignments revealed that this sequence is highly conserved (Figure 7B), and that the LMM24 mutation site is in this conserved sequence. kinase domain ( Figure 7A); 108-393 amino acids encode the kinase domain. Multiple series of alignments revealed that this sequence is highly conserved (Figure 7B), and that the LMM24 mutation site is in this conserved sequence.

Subcellular Localization of OsRLCK109
To determine the subcellular localization of OsRLCK109, we fused OsRLCK109 with GFP, under the control of the constitutive cauliflower mosaic virus 35S promoter, to obtain the construct 35S::RLCK109-GFP. Then, we transformed the 35S::RLCK109-GFP plasmid into rice protoplasts and co-transformed an 35S::Ghd7.1-CFP plasmid expressed in the nucleus as a marker [35]. We observed the fluorescent signals under a confocal microscope and found green fluorescent signals in both cytoplasm and nucleus where the signal in the nucleus was stronger than in the cytoplasm. This stronger GFP signal was co-localized with the nucleus marker Ghd7.1-CFP (Figure 8). These results suggest that the OsRLCK109 protein was localized mainly in the nucleus but there is also a weaker signal in the cytoplasm.

Subcellular Localization of OsRLCK109
To determine the subcellular localization of OsRLCK109, we fused OsRLCK109 with GFP, under the control of the constitutive cauliflower mosaic virus 35S promoter, to obtain the construct 35S::RLCK109-GFP. Then, we transformed the 35S::RLCK109-GFP plasmid into rice protoplasts and co-transformed an 35S::Ghd7.1-CFP plasmid expressed in the nucleus as a marker [35]. We observed the fluorescent signals under a confocal microscope and found green fluorescent signals in both cytoplasm and nucleus where the signal in the nucleus was stronger than in the cytoplasm. This stronger GFP signal was co-localized with the nucleus marker Ghd7.1-CFP (Figure 8). These results suggest that the OsRLCK109 protein was localized mainly in the nucleus but there is also a weaker signal in the cytoplasm.
co-transformed an 35S::Ghd7.1-CFP plasmid expressed in the nucleus as a marker [35]. We observed the fluorescent signals under a confocal microscope and found green fluorescent signals in both cytoplasm and nucleus where the signal in the nucleus was stronger than in the cytoplasm. This stronger GFP signal was co-localized with the nucleus marker Ghd7.1-CFP (Figure 8). These results suggest that the OsRLCK109 protein was localized mainly in the nucleus but there is also a weaker signal in the cytoplasm.

LMM24 is the OsRLCK109 Gene and Functional Mutation of OsRLCK109 Leads to the Mutant Phenotype
Many LMMs have been identified in rice and have been widely applied to the study of PCD and defense responses. In this study, we isolated an LMM, lmm24, from the mutant pool of indica rice "ZhongHui8015". Dark brown lesions spontaneously formed in lmm24. Moreover, compared with ZH8015, mutant plants were shorter and had fewer tillers, a later growth period, and lower yield (Figure 1 and Figure S1). Like most LMMs, the lesions mainly appeared on the leaves and the growth of the plant was affected; however, the lesions of lmm24 differed from those of most LMMs because production of lesions were light-dependent ( Figure 1). The LMM24 gene was identified as LOC_Os03g24930 by MutMap, which encodes OsRLCK109. Sequencing results showed that the replacement of two single bases and a 54 bp insertion occurred on the fourth exon of LMM24, resulting in the change of two amino acids in the kinase domain of OsRLCK109. Mutagenesis by EMS often results in single-base mutations, with fewer examples of large fragment insertions and single-base mutations. The functional complementation of OsRLCK109 restored the mutant to the wild-type phenotype, indicating that the functional mutation of OsRLCK109 is responsible for the mutant phenotypes.

Cell Death and Early Leaf Senescence are Caused by Functional Mutation of OsRLCK109
The lmm24 mutant displayed spontaneous cell death from the seedling stage to the yellow mature stage. Experimental results show accumulation of H 2 O 2 in lmm24. H 2 O 2 is a major by-product of beta-oxidation and acts as a signal molecule in the promotion of cell death, thus, excessive accumulation of H 2 O 2 is the main cause of lesion formation. Like most LMMs, lmm24 accumulates H 2 O 2 only on leaves. Unfortunately, more research is needed to find the cause. Early leaf senescence also occurred in lmm24 and we found chloroplast degradation in lesion areas of lmm24 by TEM as well as down-regulation of photosynthesis-related genes and up-regulation of the SGR gene and senescence-associated genes. In previous studies on early leaf senescence, senescence and lesions were described as two different traits; only a few studies have described early leaf senescence as one of the characteristics of LMMs [6,9]. We believe that lesions are typical of early leaf senescence and that leaf senescence in LMMs usually happens rapidly, providing a unique tool for understanding leaf senescence.

OsRLCK109 Regulates Defense Response in Rice from Disease
The defense response is activated in LMMs to increase broad-spectrum disease resistance to pathogens. We found that lmm24 improved resistance to rice blast by inoculation experiments and that PR genes were all significantly upregulated. These results suggest that the defense response is activated in lmm24, but the molecular mechanism of this resistance is unclear. LMM24 encodes OsRLCK109 and, reportedly, RLCKs play a key role in the transmission of immune signals, by associating with immune RKs. Pattern recognition receptors (PRRs) located on the plasma membrane are required to perceive the conserved pathogen-associated molecular patterns (PAMPs) during plant innate immunity. Chitin elicitor receptor kinase 1 (OsCERK1) is a type of PRR; the RLCK family members OsRLCK185, OsRLCK176, OsRLCK57, OsRLCK107, and OsRLCK118 were all reported to be located at the plasma membrane because of their interactions with OsCERK1. However, in our study, OsRLCK109 was located in the nucleus and cytoplasm, so we speculated that OsRLCK109 is involved in an immune pathway different from these RLCKs, but further experimental evidence is needed to support our inference.
In conclusion, functional mutation of OsRLCK109 induces spontaneous lesions in leaves and enhances defense responses in lmm24. Our results suggest that OsRLCK109 may play an important role in regulating cell death and defense responses in rice. Recent advances suggest a key role of RLCKs in RK-mediated signaling. Correspondingly, most RLCKs are localized to the cell membrane. In this study, RLCK109 was mainly located in the nucleus, which differs from findings of previous studies and may provide new avenues to further investigate the role of RLCKs in the immune signaling pathway.

Plant Material and Growth Conditions
The lmm24 mutant was isolated from a mutant pool of indica rice ZH8015 by ethyl methane sulfonate (EMS) treatment. A cross was made between lmm24 and ZH8015, the first filial generation (F 1 ) was self-pollinated, and the second generation (F 2 ) was used for genetic analysis. The lmm24 and ZH8015 seeds were grown in a plant growth chamber (14 h light/10 h dark, 28/25 • C, 90% relative humidity) for an inoculation experiment with rice blast. All F 2 and parent populations were grown in an experimental paddy field at the China National Rice Research Institute (Hangzhou, Zhejiang province in China) from May to November, 2017.

Mapping of LMM24
MutMap was performed to map the LMM24 gene according to a previous study [29]. We generated an F 2 population from a cross between lmm24 and ZH8015. DNA from the 30 plants with lmm24 phenotypes from this F 2 population was pooled using the same amount of DNA from each plant for whole genome sequencing. DNA purity and integrity were analyzed by agarose gel electrophoresis, and DNA purity was determined by Nanodrop2000 (Thermo Fisher Scientific, Waltham, MA, USA). The qualified DNA samples were then randomly broken into fragments of 350 bp in length by a Covaris crusher. The library was built using the TruSeq Library Construction Kit (Illumina, San Diego, CA, USA), and the constructed library was sequenced by an Illumina HiSeqTM PE150 (Illumina, San Diego, CA, USA). The indica 93-11 (http://www.rise.genomics.org.cn) genome sequence was used as a reference genome sequence and the SNP/InDel index was calculated according to a previous report [36]. cDNA of LOC_Os03g24930 with a two SNP index of 1 was PCR amplified using primers LMM24F/R. The product was subsequently sequenced to confirm the mutation.

Complementation of LMM24
A 5192 bp genomic fragment of ZH8015 was PCR amplified using primer pairs P-LMM24 for complementation of the lmm24 mutant phenotype. This fragment consisted of an 1842 bp upstream promoter (Table S2), a 2184 bp gene region, and a 1166 bp downstream terminator. The PCR product was recombined in the binary vector pCAMBIA1300 using an In-Fusion Advantage Cloning kit (Clontech, San Francisco, CA, USA). The resultant expression construct was transformed into lmm24 by Agrobacterium tumefaciens-mediated transformation. The transformation experiment was completed by Wuhan Boyuan Biotechnology Co., Ltd (Wuhan, China).

Histochemical Marker Staining Assay
Leaves (second from the top) of field-grown lmm24 and ZH8015 at 50 dps were used in this assay. Evans blue staining for dead cells and DAB staining for H 2 O 2 accumulation were conducted as previously described [37,38].

TEM and TUNEL Assay
Samples of ZH8015 and lmm24 leaves (second from the top) at 50 dps were prepared for TEM. The leaf segments were fixed in a 2.5% glutaraldehyde solution at 4 • C overnight. The samples were rinsed three times, each time for 15 min, with 0.1 M, phosphate buffer at PH 7.0. The samples were then fixed with 1% osmic acid solution for 1-2 h, dehydrated in a graded ethanol series and embedded in Spurr's resin. Ultrathin samples were made as previously described [39,40]. In order to identify nuclear DNA fragmentation, which is also a distinct feature of PCD, we used the TUNEL assay; the test method is based on the report of Ku et al. [41].

RNA Isolation and qPCR Analysis
Total RNA was isolated from leaves (second from the top) of ZH8015 and lmm24 at 50 dps with the TIANGEN RNAprep Pure Plant Kit (TIANGEN, Beijing, China). cDNA was synthesized with a ReverTra Ace kit (TOYOBO, Osaka, Japan) and qPCR was performed on a Roche LightCycler 480 (Roche, Basel, Switzerland) device using a SYBRPremix ExTaq kit (Takara, Dalian, China). The ratio of A260/A280 was 1.8-2.2, and two clear bands of 18s RNA and 28s RNA were distinguishable by agarose gel electrophoresis. Three biological replicates and three technical replicates were performed for each sample. The expression level of the target genes was normalized to that of the β-actin1 gene and Ubiquitin gene (Table S5). The results of qPCR were statistically analyzed using the 2 -∆∆Ct method. In addition, the values of ZH8015 were normalized to clearly show the change of lmm24 relative to ZH8015.
Plant materials were grown in the experimental field of the China Rice Research Institute. The leaves were immediately immersed in liquid nitrogen and brought back to the laboratory for storage at −80 • C. DNA was removed using DNase for RNA extraction and DNA was removed again using the ReverTra qPCR gDNA Remover kit before reverse transcription. The cDNA was diluted twice.

Subcellular Localization
The full-length CDS of LMM24 was amplified by GFP-F/R and the PCR product was cloned into the C-terminus of the green fluorescent protein (GFP) coding region in pJIT163-hGFP vectors. Plasmids 35S::RLCK109-GFP and 35S::GFP were transformed into rice protoplasts prepared from leaf tissues using the polyethylene glycol (PEG)-mediated transformation method. A 35S::Ghd7.1-CFP plasmid was co-transformed as a marker. Fluorescence of GFP was detected using a laser confocal scanning microscope ZEISSLSM 700 (ZEISS, Jena, Germany) 48 h after transfection. GFP can be excited by the 488 nm laser line and is optimally detected at 600 nm. ECFP (enhanced cyan fluorescence protein) was excited at 405 nm and was optimally detected at 550 nm. Mercury lamp was used to excite the fluorescent proteins.

Pathogen Infection
For disease resistance evaluation, leaves of ZH8015 and lmm24 were inoculated with virulent M. oryzae pathotypes following a previously described procedure [42].