Lotus japonicus belongs to the subfamily Papilionoideae, being named after its unique papilionaceous (butterfly-like) flowers, and provides an excellent experimental system to investigate the molecular mechanisms in the control of zygomorphic flower development. L. japonicus has been adopted as a model legume plant in genomic and the molecular genetic study in the late 1980s, being recognized for its relative simple genome structure, small genome size, feasible for genetic transformation and other unique characteristics (Arumuganathan et al., 1991; Asamizu et al., 2000; Blondon et al., 1994; Greilhuber et al., 1997;Handberg and Stougaard, 1992; Kawasaki and Murakami, 2000; Sandal et al., 2002). 

In recent years, there has been a great progress in legume genomics due to the genome sequencing program in a few model legume plants, including the one of L. japonicus (Cannon et al, 2005; Grant et al, 2009; Lamblin et al, 2003; Sato et al, 2008; Shoemaker et al, 2008)  Nevertheless, screening mutants is still an efficient research strategy to identify the key genes in the regulatory network controlling important biological process. In the previous study, we have reported several important loci being found in the mutagenesis and a few important key regulators in the control of floral development were cloned from our mutagenized population of L. japonicus (Dong et al, 2005; Chen et al, 2006; Feng et al, 2006).  

L. japonicus possesses the flowers which are zygomorphy and display the dorsoventral (DV) asymmetry: there is a DV axis in their floral plane, along which there exist a dorsal, two lateral and two ventral petals; three types of petals are different in shape and size (Figure 1 a). In the previous study, a mutation, keeled wing in Lotus 1 (kew1), was analyzed (Figure 1 b; Feng et al, 2006), which affects the development of lateral petal and is named for its similar phenotype to the mutant keeled wing (k) in pea (Pellew and Sverdrup, 1923; Ellis et al, 1992). Mutant flowers of both kew1 and k bear the aberrant lateral petals which acquire the same identity as the one of ventral petal in the wild type (Figure 1 b). Comparative mapping experiment demonstrated that KEW1 and K should be the orthologous regulators and play an important role in the control of lateral petal development in papilionoid legumes (Feng et al, 2006). Recently, K has been cloned and is found to encode a TCP transcriptional factor (Wang et al, 2008). In L. japonicus, LjCYC3, the homolog of K, was isolated but no detectable mutation has been identified in a 5 kb region containing the LjCYC3 gene in the kew1 mutant (Wang et al., 2008).

In this study, we reported our effort to conduct large scale mutagenesis by using physical-mutagenesis (γ-ray) and chemical-mutagenesis (EMS) in L. japonicus. Two new mutants, kew2 and kew3, possessing disrupted shape or identity of lateral petal were isolated and characterized. Genetic analysis indicated that kew3 was caused by single locus and allelic to the kew1. However, kew2 was a variable mutant and display weaker malfunction on the shape of lateral petals than kew1 or kew3, suggesting another genetic locus might have involved in the lateral petal development. With more mutants and genetic loci were identified, the function of the key regulators in the control of lateral petal development will be elucidated in the coming future.

1 Results
1.1 Physical-mutagenesis and chemical-mutagenesis in Lotus japonicus
Physical-mutagenesis (γ-ray) and chemical-mutagenesis (EMS) were used to construct the mutation library in L. japonicus, respectively. Different dosages of mutagenesis were evaluated by testing the germination rates, and 50% germination rate was set as the limit dosage for the treatment. Then, 25 000 seeds of ecotype Gifu were treated with γ-ray (250 Gy) or EMS (0.75%) respectively. The seeds were germinated and sown, giving rise to about 20 000 plants and 30 000 plants in two M1 populations in two growing seasons respectively. Seeds were collected from each M1 plant and packed separately. 13000 and 25000 M2 families were grown in two separate growing periods in 2002 and 2005 respectively. Individual plants of the M2 generations were screened for development mutations. Several floral mutations and leaf mutations were obtained and selected for detailed analysis (data not shown). 

1.2 Mutations possessing aberrant lateral petals 
Two mutants involved in the malfunction of lateral petal development were obtained in the γ-ray mutagenized M2 population. These mutants displayed no other detectable phenotype apart from their common characteristic: the lateral petals have abnormal shapes, while the petals in dorsal and the ventral positions are as normal as ones of the wild type. Since their phenotypes mimic the one of kew1, they were designated as kew2 and kew3 respectively (Figure 1c;Figure 1d).

When the shapes of lateral petals were compared among kew1, kew2 and kew3, the subtle difference between kew1 and kew3 could be seen: there are wrinkles at the edge of the kew3’s lateral petal; however, the shape of kew2’s lateral petal is quite different from the one of either kew1/kew3 or the wild type. It is evident that the identity of lateral petals of kew1 and kew3 is altered, which mimics the one of ventral petal in the wild type (Figure 1a, Figure 1b and Figure 1d). In a comparison with the wild type, the lateral petal of kew2 is narrower and possesses a tip at the top end (Figure 1c).
 

1.3 Genetic analysis of kew2 and kew3
When kew3 was back-crossed with the wild type parent line Gifu B129, all flowers of F₁ are wild type, while F₂ progenitors gave wild type and mutant flowers in 3:1 ratio (n>1500), indicating kew3 is caused by a single recessive locus. After several backcrosses with the wild type in which 3:1 ratios of the wild type and mutant were observed consistently in the segregating populations, kew3 behaved as a stable mutant. However, in the backcross with the wild type, the phenotype of kew2 could disappear in the following generations.

Since the phenotype of kew3 mimicked the one of kew1, allelic testing was conducted. It was found that kew1 and kew3 were allelic, since all the F₁ progenies displayed the mutant phenotype as the one of kew1 and kew3. However, in the allelic test of kew2, the F₁ progenies of the cross between kew1 and kew2 were segregated as either the wild type or kew1. These suggest that in the kew2 genetic background, the kew1 locus must be mutated but should be in a heterozygous state, suggesting that other genetic locus/loci were contributing to the mutant phenotype in the original kew2 mutant. 

2 Discussion
Floral zygomorphy (flower with bilateral symmetry), as a specialized forms of flower symmetry, is an important evolutionary trait in flowering plants with multiple origins. The molecular analysis on the mechanism in the control of zygomorphic development was first conducted in a model plant, Antirrhinum majus, that was commenced from the screening of floral mutants from mutagenesis experiments  (Carpenter and Coen, 1990). Combination of precise genetic analyses and transposon-tagging approach, a few key regulatory genes were identified and cloned. In the past ten years, we have explored the model papilionoid legume, Lotus japonicus, which has the advantages for molecular genetic studies due to its relative smaller size, simple genome structure, feasible for genetic transformation and other characteristics. In this study, we reported our effort to conduct large scale mutagenesis in L. japonicus. 

Ionizing radiation, such as γ and X–ray, are the efficient physical mutagens, which produce deletions leading to chromosome break and chromosome aberrations. On the other hand, most chemical mutagens normally cause base pair substitutions, especially from GC to AT in the case of EMS. With the difference in the mechanism, the different mutagenesis, such as, physical and chemical mutagens have different hotspots, produced different spectrum (Van der Veen. 1966; Feldmann, 1991). In our mutagenesis in L. japonicus, a few kew mutants were identified in the γ-ray mutagenized M2, but EMS did not produce the kew mutation yet in our experiment. This could either be a by-chance case or alternatively indicates that different type of mutations might be obtained through different mutagenesis. 

In Antirrhinum majus, two TCP genes, CYCLOIDEA (CYC) and DICHOTOMA (DICH), and two MYB genes, RADIALIS (RAD) and DIVARICATA (DIV) play key role in the development of zygomorphic flower, and their interplay determine the petal differentiation (Carpenter and Coen 1990; Luo et al, 1996; 1999; Almeida et al, 1997;Galego and Almeida, 2002). It has been shown that both CYC and DICH control the dorsal and lateral petal identity during zygomorphic development. In cyc dich double mutant, all petals resemble the ventral petals of wild type. RAD and DIV interplay with CYC and DICH, and two pairs of key regulators control the asymmetry of the whole flower, as well as the one of lateral and dorsal petals in Antirrhinum majus (Corley et al, 2005).

L. japonicus belongs to Papilionoideae, whose floral asymmetry is thought to have an independent origin, different from the one in Antirrhinum majus. In our recent study on the floral zygomorphy in pea, another well known model plant in papilionoid legumes, we identified two kinds of key regulatory genes in the control of zygomorphic flower development, and proposed that they determine the two kinds of asymmetries independently: the dorsoventral (DV) and organ internal (IN) asymmetries in the floral plane and organ plane respectively (Wang et al, 2008). K and KEW play the important role in the control of lateral petal development and thus are the key regulatory genes in the DV pathway. As a comparison, no single locus which controls the lateral petal development has been identified in Antirrhinum majus. 

In this study, we characterized two kew mutants. Genetic analysis indicated that they were allelic to kew1, a mutant locus being identified as the homolog one of k based on the comparative genomic data in the previous study (Feng et al, 2006). K has been cloned and encodes a TCP factor PsCYC3 (Wang et al, 2008). The most closely homolog of PsCYC3 in L. japonicus is LjCYC3. However, in the genomic sequence of LjCYC3, no mutation has been identified yet. With more alleles of kew1 were isolated, the molecular base for the kew mutation should be addressed and solved soon. Furthermore, genetic analysis of kew2 indicated that the original kew2 mutant should be heterozygous at the kew1 locus. In the analysis of kew1 and kew3, it has been shown that kew mutant is single recessive one. These data suggest that there are other genetic factors in the genome, which could interact with kew locus and determine the lateral petal development. Cloning and functional analysis of K/KEW and other unique loci in legumes genome should provide a good chance to dissect the molecular base on the evolution diversity between different zygomorphic developments.

3 Materials and Method
Lotus japonicus (Gifu B-129) was used in this experiment. Before planting, the seeds were imbibed in liquid N2 for 1min to break the tough capsule. M1 Plants were grown in greenhouse and maintained under 16h light, 8h darkness using supplemental light. M2 plants, 20~30 plants for each family, were grown in fields, Zhonggu Farm, Shanghai.

Acknowledgements
We thank Xudong Zhu for γ-ray irradiation and this work is supported by a funding (2009ZX08009-112B) from the Ministry of Agriculture, China. 

References 
Almeida J., Rocheta M., and Galego L., 1997, Genetic control of flower shape in Antirrhinum majus, Development, 124: 1387-1392

Arumuganathan K., and Earle D., 1991, Nuclear DNA content of some important plant species, Plant Molec. Biol. Reporter, 9: 208-218 doi:10.1007/BF02672069

Asamizu E., Watanabe M., and Tabata S., 2000, Large Scale Structural Analysis of cDNAs in the Model Legume, Lotus japonicus, J. Plant Res., 113: 451-455 doi:10.1007/PL00013955

Blondon F., Marie D., Brown S., and Kondrosi A., 1994, Genome size and base composition in Medicago sativa and M. truncatula species, Genome, 37: 264-270 doi:10.1139/g94-037

Cannon S.B., Crow J.A., Heuer M.L., Wang X., Cannon E.K.S., Dwan C., Lamblin A.F., Vasdewani J., Mudge J., Cook A., Gish J., Cheung F., Kenton S., Kunau T.M., Brown D., May G.D., Kim D., Cook D.R., Roe B.A., Town C.D., Young N.D., and Retzel E.F., 2005, Databases and Information Integration for the Medicago truncatula Genome and Transcriptome, Plant Physiol., 138(1): 38 - 46 doi:10.1104/pp.104.059204

Carpenter R., and Coen E., 1990, Floral homeotic mutations produced by transposon-mutagenesis in Antirrhinum majus, Genes Devel., 4(9): 1483-1493 doi:10.1101/gad.4.9.1483

Chen J.H., Pang J.L., Wang L.L., Luo Y.H., Li X., Cao X.L., Lin K., Ma W., Hu X.H., and Luo D., 2006, Wrinkled petals and stamens 1, is required for the morphogenesis of petals and stamens in Lotus japonicus, Cell Research, (16): 499-506 doi:10.1038/sj.cr.7310061

Corley S.B., Carpenter R., Copsey L., and Coen E., 2005, Floral asymmetry involves an interplay between TCP and MYB transcription factors in Antirrhinum, Proc. Natl. Acad. Sci. USA., 102, 5068–5073 doi:10.1073/pnas.0501340102

Dong Z.C., Zhao Z., Liu C.W., Luo J.H., Yang J., Huang W.H., Hu X.H., Wang T.L., and Luo D., 2005, Floral patterning in Lotus japonicus, Plant Physiol., 137:1271-1282 doi:10.1104/pp.104.054288

Ellis T.H.N., Turner L., Hellens R.P., Lee D., Harker C. L., Enard C., Domoney C., and Davies D.R., 1992, Linkage Maps in Pea, Genetics, 130: 649-663

Feldmann K.A., 1991, T-DNA insertion mutagenesis in Arabidopsis: mutaional spectrum, Plant J., 1: 71-82 doi:10.1111/j.1365-313X.1991.00071.x

Feng X.Z., Zhao Z., Tian Z.X., Xu S.L., Luo Y.H., Cai Z.G., Wang Y.M., Yang J., Wang Z., Weng L., Chen J.H., Zheng L.Y., Guo X.Z., Luo J.H., Sato S., Tabata S., Ma W., Cao X.L., Hu X.H., Sun C.R., and Luo D., 2006, From the Cover: Control of petal shape and floral zygomorphy in Lotus japonicus, Proc. Natl. Acad. Sci. USA., 103:4970-4975 doi:10.1073/pnas.0600681103

Galego L., and Almeida J., 2002, Role of DIVARICATA in the control of dorsoventral asymmetry in Antirrhinum flowers, Genes Dev., 16(7): 880-891 doi:10.1101/gad.221002

Grant D., Nelson R.T., Cannon S.B. and Shoemaker R.C., SoyBase, the USDA-ARS soybean genetics and genomics database, 2009, Nucleic Acids Research Advance Access originally published online on December 14, 2009(D843-D846; doi:10.1093/nar/gkp798)

Greilhuber J., and Oberrmayer R., 1997, Genome size and maturity group in Glycine max (soybean), Heredity, 78: 547-551 doi:10.1038/hdy.1997.85

Handberg K., and Stougaard J., 1992, Lotus japonicus, an autogamous, diploid legume species for classical and molecular genetics, Plant J., 2: 487-496 doi:10.1111/j.1365-313X.1992.00487.x

Kawasaki S., and Murakami Y., 2000, Genome analysis of Lotus japonicus, J. Plant Res., 113: 497-506 doi:10.1007/PL00013960

Lamblin A.F.J., Crow J.A., Johnson J.E., Silverstein K.A., Kunau T.M., Kilian A., Benz D., Stromvik M., Endré G., VandenBosch K.A., Cook D.R., Young N.D., and Retzel E.F., 2003, MtDB: a database for personalized data mining of the model legume Medicago truncatula transcriptome, Nucleic Acids Res., 31(1): 196-201 doi:10.1093/nar/gkg119

Luo D., Carpenter R., Vincent C., Copsey L., and Coen E., 1996, Origin of floral asymmetry in Antirrhinum, Nature, 383: 794-799 doi:10.1038/383794a0

Luo D., Carpenter R., Copsey L., Vincent C., Clark J., and Coen E., 1999, Control of organ asymmetry in flowers of Antirrhinum, Cell, 99(4): 367-376

Pellew C., and Sverdrup A., 1923, New observations on the genetics of peas, Journal of Genetics, 13: 125-131 doi:10.1007/BF02983043

Sandal N., Krusell L., Radutoiu S., Olbryt M., Pedrosa A., Stracke S., Sato S., Kato T., Tabata S., Parniske M., Bachmair A., Ketelsen T., and Stougaard J.A., 2002, Genetic Linkage Map of the Model Legume Lotus japonicus and Strategies for Fast Mapping of New Loci, Genetics, 161(4): 1673-1683

Sato S., Nakamura Y., Kaneko T., Asamizu E., Kato T., Nakao M., Sasamoto S., Watanabe A., Ono A., Kawashima K., Fujishiro T., Katoh M., Kohara M., Kishida Y., Minami C., Nakayama S., Nakazaki N., Shimizu Y., Shinpo S., Takahashi C., Wada T., Yamada M., Ohmido N., Hayashi M., Fukui K., Baba T., Nakamichi T., Mori H., and Tabata S., 2008, Genome structure of the legume, Lotus japonicus, DNA Res., 15: 227-239 doi:10.1093/dnares/dsn008

Shoemaker R.C., Grant D., Olson T., Warren W.C., Wing R., Yu Y., Kim H., Cregan P., Joseph B., Futrell-Griggs M., Nelson W., Davito J., Walker J., Wallis J., Kremitski C., Scheer D., Clifton S.W., Graves T., Nguyen H., Wu X., Luo M., Dvorak J., Nelson R., Cannon S., Tomkins J., Schmutz J., Stacey G., and Jackson S., 2008, Microsatellite discovery from BAC end sequences and genetic mapping to anchor the soybean physical and genetic maps, Genome, 51: 294-302 doi:10.1139/G08-010

Sage G.C.M., 1995, Induced Mutations and Molecular Techniques for Crop Improvement, Proceedings of Symposium, Crop Protection, 15(6): 593-594 doi:10.1016/0261-2194(96)89794-8 

Wang Z., Luo Y.H., Li X.,Wang L.P., Xu S.L., Yang J., Weng L., Sato S., Tabata S., Ambrose M., Rameau C., Feng X., Hu X., and Luo D., 2008, Genetic control of floral zygomorphy in pea (Pisum sativum L.), Proc. Natl. Acad. Sci. USA., 105: 10414-10419 doi:10.1073/pnas.0803291105