Introduction

Genetic analysis of mutations provides a powerful tool to reveal complex processes of plant growth and development. Mutants are valuable resources that could provide genetic loci to generate improved crops. This study evaluated a rice mutant with various phenotypic expressions at the vegetative and reproductive stages, displaying lesion mimic mutant (LMM) spotted leaf 6 (spl6), liguleless (lg), and lax panicle (lax) mutant phenotypes according to phenotypic and genetic analyses.

LMM plants develop scattered dotted spots throughout leaf surfaces without stress from pathogenic infection. On the other hand, a genetically regulated signalling pathway leads to death of the localized cell, known as programmed cell death (PCD) (Balagué et al. 2003; Zeng et al. 2004; Matin et al. 2006; Mori et al. 2007; Williams and Dickman 2008; Bruggeman et al. 2015; Xia et al. 2019; Kang et al. 2021; Zheng et al. 2022; Zhao et al. 2023). Various stress elicitors, such as biotic and abiotic stresses, mechanical damage from temperature, light, wind, as well as reactive oxygen species (ROS), have been identified as influential for the PCD and lesion development in plants (Takahashi et al. 1999; Balagué et al. 2003; Matin et al. 2006; Kang et al. 2021; Wu et al. 2023). Thus far, more than 30 LMM in rice, including several LMMs in many other plants, have been studied intensively for their genetic mechanism, underscoring the importance of further research using this mutant (Yamanouchi et al. 2002; Wang et al. 2005, 2019; Ma et al. 2019; Ruan et al. 2019; Liu et al. 2021; Yuchun et al. 2021; Chen et al. 2022; Li et al. 2022; Shang et al. 2022; Wu et al. 2023). Recent extensive reviews described that most LMMs have been identified in rice and other plants (Kang et al. 2021; Yan et al. 2022). The present study analysed the spl6 mutant, which forms tiny dotted lesions at the early tillering stage in leaf and gradually increase in size.

Morphologically, the rice leaf is structured with a leaf sheath and leaf blade with a laminar joint between them called the collar, surrounded by two hairy sickle-shaped auricles and a tongue-like white apparatus called the ligule. These specialized organs help to keep the leaf sheath and culm together and prevent dust entry. The collar helps to bend the leaf to facilitate photosynthesis (Hoshikawa 1989; Lee et al. 2007b). In the lg phenotype, the leaf sheath is connected directly to the leaf blades without a collar region. Therefore, the ligule and auricle of lg are not formed resulting in upright and high erectness of leaf. Similarly, rice lg mutant, OsLG1, have been identified to lack all the three organs (Lee et al. 2007b).

In rice, the panicle density is regarded as the most important characteristic from a breeding point of view because it is closely associated with the grain yield. Rice panicles develop at the reproductive stage of the plant development by phase transition of the shoot apical meristem. After the transition, the inflorescence meristem produces primary branch meristems that develop a primary panicle branch. The primary panicle branch also produces secondary branch meristems that produce secondary panicle branches or differentiate into lateral spikelets. Finally, both meristems differentiate into a terminal spikelet (Yamagishi et al. 2004; Oikawa and Kyozuka 2009). Thus, the pattern of apical meristem formation has particular significance in crop spikelets development.

In contrast, the lax panicle mutant is characterized by a few elongated rachis branches and reduced spikelets on the panicle. Rice LAX PANICLE1 (LAX1) is required to generate the branch meristem. Hence, the lax1 mutant fails to form a branch meristem, leading to reduced spikelets (Oikawa and Kyozuka 2009; Matin and Kang 2012). The rice panicle starts via the limited growth of inflorescence consisting of primary branches, secondary branches, and spikelets on the branches. The panicle architecture affects the grain number per panicle and grain yield. Therefore, apical meristem formation and panicle structure development have attracted considerable interest for improving rice quality since they contribute directly to crop yield.

Although several phenotypic mutations have been identified so far, more intensive research on materials with clear specific phenotypic expression is needed to identify the function and regulatory mechanism of specific mutations in rice. From various individual studies, the spl6, lg, and lax genes were identified as recessive genetically controlled compared to the normal genes (Mori et al. 1973; Lee et al. 2007a,b; Matin and Kang 2012; Kang et al. 2021). On the other hand, a few studies have examined the linkage and genetic correlation of these three genes in rice. The present study used a mutant rice plant with homozygous recessive genes for these three genetic traits, and their Mendelian genetic relationships were revealed. These findings may provide fundamental information about the molecular functions of these genes in the future.

Materials and methods

Plant materials and phenotype analysis

Rice YUM111 (FL269 lines) mutant (O. sativa Japonica type) is one of the mutant lines induced by N-methyl-N-nitrosourea (MNU) that was derived at Kyushu University, Japan (Kumamaru et al. 1988). The mutant exhibited distinct spots on leaves, liguleless, and lax-type panicle phenotypes. Rice plants were grown under natural conditions between 30 and 35°C in the research field of Yeungnam University, Gyeongsan, Republic of Korea, and the phenotypic characteristics were documented. The observed characteristics were as follows: spot formation time, colour, shape, size, and arrangement of spots; the severity of spots at different developmental stages; presence and absence of a collar, auricle, and ligule at the leaf junction; panicle structure; length and spikelet formation type; number of primary rachis branches and secondary rachis branches; number of nodes per rachis branch; and number of spikelets.

Genetic analysis

For inheritance analysis, the mutant line was hybridized to wild-type pollen donors IRRI347 (YUC084) (O. sativa Indica type) and Donghechal (YUC005) (O. sativa Japonica type). The F1 plants were grown and self-fertilized to produce the F2 progeny. The phenotypic data of the F2 plants were documented throughout the developmental stages for the concerned genes. The genotype of individual F2 plants was detected from their segregating feature in the F3 progeny. The genotype of the plants was confirmed by planting 20 seeds of each dominant F2 plant in the field and determining the heterozygous plants. The phenotypic data were collected from the F3 plants, and their homozygosity and heterozygosity were detected based on their segregation. Phenotypic data were collected from 558 F3 plants. Segregation of individual gene for the monohybrid cross was evaluated based on the Mendelian 1:3 phenotypic ratios. Dihybrid crosses for the co-segregation of two genes were evaluated based on the Mendelian 9:3:3:1 phenotypic ratio. By considering crosses between the lax and spl6 genes, parental and recombinant plants were calculated, and their distance was measured in centimorgans (cM).

Results and discussion

Phenotype of the spl6 mutation

The spl6 mutant line formed spontaneous lesions on the leaf blades (Matin et al. 2006; Kang et al. 2007). The phenotypic observation indicated that no visible spots were initiated at the early stage. On the other hand, a tiny appearance occurred at the tillering stage. The lesions progressed intensively throughout the leaf surface parallel to the leaf minor vain and rarely on major vain, which was visible at the milk stage. New spots continued to appear until maturity. Lesion formation was diffused, with lesions starting sparsely and rapidly expanding over the entire leaf blade, becoming denser toward the apex and thinner toward the leaf base (figure 1).

Interestingly, flag leaves of the mutant were more heavily spotted than the other leaves. The sunlight intensity significantly affects leaf lesion formation, development, and acceleration in the mutant plants, indicating that lesion formation in the LMMs may be linked directly to the light reactions of the photosynthesis process (Yamanouchi et al. 2002; Zeng et al. 2004; Qiu et al. 2019; Zhang et al. 2019; Yao et al. 2021; Jiang et al. 2022). Moreover, the pattern of spot formation was the progressive type with the developmental stages, which is similar to those in spl3, spl4, and spl7 mutant rice (Yamanouchi et al. 2002; Matin et al. 2006; Kang et al. 2021).

Phenotype of the lg mutation

The leaves of normal rice are composed of a leaf blade and a leaf sheath separated by a laminar joint. Between the leaf blade and leaf sheath, a white band-type laminar joint, called a collar, is present from where two horn or sickle-like auricles and tongue-like ligules are overextended. The lamina joint organ connects the leaf blade to the sheath and causes the leaf blade to bend away from the vertical axis known as the leaf angle, affecting light capture in rice. Thus, this structure is a key factor for the crop yield (Liu et al. 2019; Tian et al. 2019; Wang et al. 2021; Qin et al. 2023). On the other hand, a lamina joint-like collar did not exist in the YUM111 mutant. Therefore, the ligule and auricle were also not projected, resulting in a direct assembly of the leaf blade and sheath without vertical bending. In the wild type, the leaf blade forms a long angle with the stem, whereas the mutant ligule-less leaf is erect and forms a small angle with the stem (figure 1).

Plant architecture-related genes are the critical regulators of plant growth, development, and yield (Qin et al. 2023). One of the related genes, liguleless (lg), is exclusively a key regulator of the leaf angle and flower development. Some other similar types of rice mutants that resembled the present mutant include the following: auricleless (aur), which lacks the auricle and ligule; liguleless (lg) with a defective ligule, auricle, and collar (Maekawa 1988); and collarless (col), which lacks the collar region (Sanchez 1998). Identifying relationships with leaf angle genes may help control the plant phenotypes to increase yields. Including the present genotype, most of the identified lg mutant genotypes showed erect leaves that can maximize the use of light, allow light to pass to the lower canopy, and achieved a higher yield (Wang et al. 2020, 2021; Qin et al. 2023).

Phenotype of the lax mutation

The rice panicle develops during the reproductive stage via a phase transition of the shoot apical meristem to inflorescence primordia initiation. Rice inflorescence is repeatedly branched to form primary and secondary rachis branches on the panicle axis that develop individual florets on the rachis branches and terminates in a spikelet on both primary and secondary branches that determine the panicle structure (Yamagishi et al. 2004; Bell and Bryan 2008; Yoshida et al. 2012). Dysfunction of regulatory genes at any stage of inflorescence development can result in a loose-type panicle. In rice, the lax mutant exhibits an opposite pattern in which vegetative branching is normal, but the axillary meristems are severely blocked in the reproductive stage. In YUM111 mutant rice, vegetative growth and development were normal, but after phase transition, inflorescence development was distorted, and lateral panicle branches were barely formed, resulting in fewer spikelets (figure 1).

Figure 1
figure 1

Phenotypes of the wild type (YUC005) and mutant (YUM111) plants for the spl6, lg, and lax mutations. In the mutant parent, for spl6, severely spotted matured flag leaf and magnified leaf segment; for lg, liguleless phenotype with no collar, auricle, and ligules at the leaf blade and leaf sheath junction; for lax, plant upper part and an individual panicle are demonstrated. The corresponding phenotypes of the wild-type plant are also indicated for comparison. L, ligules; C, collar; A, auricles; bl, leaf blade; sh, leaf sheath.

The mutant phenotype was evident from early inflorescence development, where loosely arranged panicle branches were observed in the mutant, while the wild type showed a compact arrangement (figure 2a). Occasionally, primary and secondary rachis branches could not develop from panicle node and primary rachis nodes, respectively, in the mutant (figure 2b). Moreover, initiation of lateral spikelets primordia was suppressed at the secondary rachis nodes, which failed to develop lateral spikelets. Furthermore, only the terminal node formed spikelets as terminal florets in the mutant, resulting in sparsely distributed and reduced seeds per panicle (figures 2a, b). The rachis internodes as well as primary and secondary rachis internodes were excessively elongated in the mutant, approximately two times longer than the wild type (figure 2c). On the other hand, the mutant had fewer rachis nodes, primary rachis branches, nodes per primary rachis branch, and secondary rachis branches per primary rachis branch than the wild type (figure 2d).

Figure 2
figure 2

Panicle phenotypic differences between the wild type and lax. (a) Developing inflorescence of the rice. The terminal spikelet (TSP) and lateral spikelets (LSP) developed in the primary and secondary rachis branch nodes of the wild type. Only TSP comes from the rachis branches nodes, being empty at the following node in the mutant. (b) Portion of the panicle at the mature stage. PD, peduncle; PN, panicle node; RA, rachis; PRB, primary rachis branch; SRB, secondary rachis branch. The dashed arrows indicate the rachis nodes and the length of the internode. The arrowheads indicate the primary rachis branch nodes. (c) Length (cm) of rachis internode (RI), primary rachis internode (PRI), and secondary rachis internode (SRI) of wild type (wt) and mutant (lax). (d) The total number of primary rachis branches (PRB) per panicle, number of nodes per primary rachis branch (N/PRB), and number of secondary rachis branches per primary rachis branch (SRB/PRB). The standard errors are from an average of five individual counts.

The lax mutations include the lax1 and lax2 alleles (Futsuhara et al. 1979; Tabuchi et al. 2011; Dai et al. 2022), where the proliferation of meristem cells begins but fails to form axillary meristems (Oikawa and Kyozuka 2009), and the significant reduction of primary and secondary branches was consistent with the present lax mutation (Komatsu et al. 2001, 2003; Matin and Kang 2012; Lv et al. 2023). Moreover, the YUM111 mutant is consistent with previous reports of mutations in the lax1-6 allele, which was confirmed to be a deletion mutant of the helix–loop–helix transcription factor, resulting in lack of lateral branch development and spikelet formation (Matin and Kang 2012).

In most of the lax and similar panicle mutants, including lax1, lax2, lax1-1, LAX2-4, and branch one seed 1-1 (bos1-1), vegetative branching was normal, but axillary meristem formation was severely suppressed in the reproductive development stage. In addition, the panicle branches failed to develop completely, and only a terminal spikelet was formed, having no lateral spikelet (Futsuhara et al. 1979; Komatsu et al. 2001, 2003; Tabuchi et al. 2011; Dai et al. 2022; Lv et al. 2023). All reports show that the final yield was reduced dramatically in lax mutations (Li et al. 2021; Dai et al. 2022; Lv et al. 2023).

Other agronomic traits of the mutant plant

In addition to the mutant phenotypes, several important morpho-physiological traits were also observed and compared with the wild-type (normal) plants using the standard evaluation system described by the IRRI. Most traits, including the number of effective tillers and panicles (TN), plant height, and 1000-grain weight (GW), were similar in the mutant plants to those of the wild-type plants. On the other hand, the days to flowering (DH) and the grain number per panicle (SN) were reduced significantly in the mutant than in the wild-type (table 1). These results were identical to the other panicle-associated mutants, which showed a lower seed-setting rate than the controls (Dai et al. 2022; Lv et al. 2023). The panicle morphology-related traits also showed significant differences. The numbers of rachis nodes (RN) and rachis branches (PBR and SBR) were significantly lower, and the rachis internodes (RIL) were elongated in the mutant compared to the wild type. In particular, only terminal spikelets formed in the mutant, which had no lateral spikelets. Therefore, the average seed per panicle was reduced drastically in the mutant (table 1).

Table 1 Important agronomic traits of wild type and mutant in rice.

Inheritance pattern of the spl6, lg, and lax genes in the segregating generation

Reciprocal cross-pollination was performed between a mutant (YUM111) and two wild-type rice lines (IRRI347 and Donghaechal). All phenotypic features were documented from the F1 and F2 plants throughout their developmental stages. The derived F1 plants did not form any lesions in the leaf until maturation. The leaf blade and leaf sheath junction point were clearly surrounded by a white collar, auricles, and ligules, indicating the wild-type phenotype for the lg character and a panicle structure similar to the wild-type plant (figures 3 and 4). Similar results were also obtained from reciprocal crosses of the parental combination. Hence, a recessive nuclear gene controlled these mutations. The phenotypic segregation of individual mutations was further investigated in F2 generation. The concerned genes were segregated into wild and mutant types (figure 3). The wild type to mutant ratio was then calculated to determine if they fit the Mendelian segregation ratio of 3:1.

Figure 3
figure 3

Representative of the phenotypic segregation at the F2 generation of the spl6, lg, and lax genes. Both parents (P1 and P2), cross-driven F1, and representatives of the F2 phenotypes of plants for the three genes of concern are shown.

Figure 4
figure 4

Phenotype of matured parental plants and their offspring at different generations. P1, mutant parent as pollen receptor (YUM111); P2, wild type parents as Indica type pollen donor (IRRI347) with a magnification of the mutant parts; F1, hybrid generation of a cross between P1 and P2; F2, representative phenotypes of the segregation populations developed from F1 selfing, where +, +, + is normal for all characters; spl6, +, + denotes the mutant for spl character and normal for other character; +, lg, + is the mutant for the lg character and normal for the other; +, + is the lax mutant for lax character and normal for other, and similar for the other.

For the segregation of the spl6 gene at a 3:1 ratio, the calculated chi-square (χ2) values of reciprocal crosses (χ23:1 =0.533 and 1.36, P=0.45 and 0.25, df=1) (table 2) proved this fitness as the values were less than the tabulated χ2 value (χ23:1 =3.84). Five hundred and fifty-eight F3 plants derived from selfing of one of the dominant F2 plants (SGK05010-71) segregated at 425 (wild type) to 133 (spl6), which also fited the Mendelian 3:1 ratio with a χ2 value of 0.405 (P=0.55, df=1), indicating the heterozygosity of the SGK05010-71 plant (table 2).

Table 2 Segregation of the spl6 gene at the F2 and F3 generations from reciprocal crosses.

Similarly, among the 124 dominant plants evaluated (table 2), 49 were homozygous dominant (no segregation), and the other 75 were heterozygous (segregated at 54:21=3:1). On the other hand, all 36 spl6 mutant plants produced only mutant progeny, indicating their recessive homozygosity. Finally, the genotypes of all of the F2 plants were also calculated to fit the Mendelian genotype segregation ratio of 1:2:1 (+/+: +/spl6: slp6/spl6) with a χ2 value of 2.71 (P=0.25, df=2) (table 3).

Table 3 Genotypic segregation ratios of the spl6 gene in the F2 populations obtained from F3 phenotypes.

In case of lg gene, the χ2 (3:1) values of segregation data from the reciprocal crosses (χ2 =0.428 and 1.84, P=0.50 and 0.15, df=1) (table 4) proved this fitness because the values were less than the tabulated χ2 value. The heterozygosity of the SGK05010-71 plant for the lg mutant was confirmed from the χ2 value of 0.464 (P=0.50, df=1) (table 4).

Table 4 Segregation of the lg gene at the F2 and F3 generations from reciprocal crosses.

In the case of the lax gene, the χ2 (3:1) results of the segregation data (1.64 and 2.77) also proved this fitness (P=0.20 and 0.10, df=1) (table 5). The heterozygosity of the SGK05010-71 plant for the lax mutant was confirmed from the χ2 value of 1.14 (P=0.30, df=1) (table 5). Overall, the spl6, lg, and lax phenotypes in rice are genetically stable and controlled by a single recessive nuclear gene.

Table 5 Segregation of the lax gene at the F2 and F3 generations from reciprocal crosses.

Majority of the identified rice spotted leaf mutants are regulated by recessive genes (Takahashi et al. 1999; Yin et al. 2000; Balagué et al. 2003; Liu et al. 2004; Zeng et al. 2004; Kang et al. 2021; Yuchun et al. 2021; Yan et al. 2022). On the other hand, with some exceptions, Spl7, Spl18, and Spl36 are dominant (Yamanouchi et al. 2002; Mori et al. 2007; Cai et al. 2021). In wheat, the TaSpl1 of the spotted leaf mutant is controlled by a dominant gene, which was repressed by two other dominant genes, TaSpl1-I1 and TaSpl1-I2 (Zhang et al. 2021).

The lg and lax mutations are also controlled by a single recessive gene. The lg gene has been reported to be recessive located on rice chromosome 4 (Nagao and Takahashi 1963; Lee et al. 2007b;). Rice liguleless mutant, Oslg1-3 was inherited as a single recessive gene model, and the corresponding gene OsLG1 was localized to rice chromosome 4 (Morinaga 1938; Maekawa et al. 1981; Xu et al. 2010). Becraft et al. (1990) reported one of the important discoveries in plant morphology and development. They found that, the maize liguleless-1 gene affects ligule development and is controlled by a recessive mutation. Further studies reported that single recessive genes also control maize liguleless genes lg2 and lg3 (Harper and Freeling 1996; Muehlbauer et al. 1997). Furthermore, some dominant Lg mutations have also been identified in maize (Fowler and Freeling 1996). On the other hand, most of the rice lax alleles are recessive and identified on chromosome 1 (Kinoshita 1995; Nagato and Yoshimura 1998; Komatsu et al. 2001, 2003; Matin and Kang 2012). By contrast, the rice lax2 and lax2-4 mutations were mapped on chromosome 4 controlled by a single nuclear recessive gene, Os04g0396500 (Tabuchi et al. 2011) and LOC_Os04g32510 (Dai et al. 2022), respectively. The maize lax gene has also been identified as recessive (McSteen et al. 2000).

Linkage analysis among spl6, lg, and lax genes

The YUM111 genotype exhibited spl6, lg, and lax gene mutations, and the F1 plants had normal phenotypes similar to the wild-type parent regarding all three mutations, indicating the recessive nature of each gene. When considering the co-segregation of spl6 and lg genes, they segregate independently and fit the Mendelian 9:3:3:1 ratio. The χ2 result of the segregation data (χ2=6.19, P=0.10, df=3) (table 6) proved this fitness because the value was less than the χ2 table value (χ2 9:3:3:1 =7.815).

The co-segregation of the lg and lax genes also segregated independently of each other according to the Mendelian 9:3:3:1 ratio because the calculated χ2 value (χ2=6.35, P=0.08, df=3) (table 6) was less than the tabulated χ2 value (χ2 9:3:3:1 =7.815). Hence, the spl6 and lg genes; and the lg and lax genes are not linked.

On the other hand, co-segregation of spl6 and lax genes did not follow the independent segregation pattern of the Mendelian 9:3:3:1 ratio. The calculated χ2 value (χ2=363.67 P>0.0001, df=3) was much higher than the tabulated value of χ2 29:3:3:1 =7.815), which rejects the hypothesis of their independent segregation (table 6). Therefore, spl6 and lax genes are linked to each other.

Table 6 Co-segregation of the spl6 and lg, spl6 and lax, and lg and lax genes at the F3 generations from dihybrid cross.

With three genes not linked to Mendelian genetics, each parent can produce eight different gametes. This produces 64 possible genotype combinations, and the phenotypic ratio of the trihybrid cross is 27:9:9:9:3:3:3:1, which is the visible trait of the offspring at the end of the trihybrid cross. In the current experiments, eight different phenotypic combinations of three genes were obtained in segregating generations. These include the wild type for all genes (+, +, +), the mutant for spl6 and the other wild type (spl6, +, +), the mutant for lg and the other wild type (+, lg, +), the mutant for lax and the other wild type (+, +, lax), the mutant for spl6 and lg (spl6, lg, +), the mutant for lg and lax (+, lg, lax), the mutant for spl6 and lax (spl6, +, lax), and all three mutants (spl6, lax, lg) (figure 4). On the other hand, these phenotypic combinations were not in accordance with the Mendelian trihybrid segregation that also approved the linkage of at least any gene pair.

Crossing two plants with different genotypes will result in offspring with different genotypes compared to parents. Offspring with different genotypes from their parents are the recombinant type. The recombination percentages and their standard errors are calculated if a significant deviation is detected. A significant deviation was detected for spl6 and lax segregations. Therefore, the recombination percentage between them was calculated. In the case of spl6 and lax segregation, out of 548 segregating individuals, 372 were dominant, and 124 were recessive parental types, whereas 20 and 32 plants were the recombinant type, indicating 90.51% are the parental classes, and 9.48% are the recombinant classes (table 6). Hence, the two genes are linked because the number of parent classes was greater than the recombinant class. The linkage between spl6 and lax was estimated to be in the repulsion phase from the calculated recombination percentages because of excess of lax plants in F2 populations. Therefore, the frequency of a single cross-over between them was calculated to be ~9.4%, indicating spl6 and lax might be 9.4 cM apart from each other.

Construction of linkage maps is essential for genetics, molecular research, and breeding programmes, especially for genotypes with multiple mutation phenotypes. Here, the lg gene was segregated independently with spl6 and lax at 9:3:3:1, indicating that they are not linked individually. On the other hand, the lax gene did not follow the 9:3:3:1 ratio but was co-segregated with spl6, indicating that they are linked in chromosome 1. In addition, in this study, lax and spl6 were located 9.4 cM away from each other. In the rice physical map, lax is located at the 121.1 cM locus, and spl6 is located at the 126.6 cM locus (figure 5) (Kishimoto et al. 1992; Kinoshita, 1995; Yoshimura et al. 1997).

Figure 5
figure 5

Modified from Yoshimura et al. (1997).

Linkage map of the rice chromosomes. The genes and their corresponding location in cM are demonstrated for chromosomes 1 and 4.

Conclusion

Phenotypic mutations are an excellent source for exploring the genetic mechanisms responsible for specific mutations and accurately detecting mutation points. Therefore, the mutant genotype containing three mutations, spl6, lax, and lg, provides the opportunity to reveal multiple mechanisms and common effects in a single experiment. LMMs are ideal for studying disease resistance and PCD in plants to improve crop yields. Therefore, spl6 LMM provides a valuable tool to reveal the molecular mechanisms determining PCD in rice. The lg and lax genes can describe the plant morphology and architecture. These two genes have tillering and panicle branching as important agronomic traits associated with the rice grain yield. Therefore, identifying lg and lax genes and elucidating their molecular mechanisms will be useful for rice production management and genetic improvement. The phenotype of each mutation in these materials was studied intensively. The results showed that they are controlled in a recessive manner. Therefore, understanding lax genes, which determine the rachis length, branching, rachis node density, and inter-rachis length of spikelet development, may contribute to breeding for higher yielding rice.