Isolation of a novel root-determined hypernodulation mutant rdh1 of Lotus japonicus


: H. OYAIZU, Biotechnology Research Center, University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan. Email:


A mutant line that develops an excess number of small nodules was found in Lotus japonicus Miyakojima MG20 during a screening for mutants defective in nodule development and nitrogen fixation. Genetic analysis revealed that the phenotype is inherited in a monogenic, recessive manner. The gene's locus was mapped on chromosome 1 between 53.7 and 61.4 cM. This mutant formed 5–10-fold more nodules than the wild-type plant, and a grafting experiment revealed that the root regulated the hypernodulation. Except for the nodulation's phenotype no other differences were found between the mutant and the wild-type plant with respect to growth and morphological characteristics. In the mapped locus for the mutant no nodulation genes were reported, and this fact strongly suggests that the gene's locus is a new one. The gene was named root-determined hypernodulation (rdh) 1.


Leguminous plants form nitrogen fixing root nodules with symbiotic bacteria called rhizobia. This symbiotic system for nitrogen fixation enables the legumes to grow in soil with limited nitrogen supply. The number of nodules is comprehensively controlled to provide maximal benefits to the host plant, which does not tolerate an excess number of nodules as it deprives the host cell of nutrients and photosynthetic products. Furthermore, the host plant does not appreciate nodule formation when sufficient nitrogen is available in the roots. The regulation of the number of nodules is termed “autoregulation of nodulation”, and it includes systematic suppression of earlier nodulation events (Caetanoanolles and Gresshoff 1991; Oka-Kira and Kawaguchi 2006). The mechanism of autoregulation was studied intensively for the mutant nts1 of the soybean, which forms considerably more nodules than the wild-type plant (Carroll et al. 1985). Grafting and split-root experiments showed that autoregulating signals are produced in shoots and transferred to the roots (Delves et al. 1986; Hamaguchi et al. 1993; Olsson et al. 1989; Sheng and Harper 1997). In peas, a root-regulated mutant, nod3, was found (Sagan and Duc 1996), while in the model legume Lotus japonicus and Medicago truncatula, similar hypernodulation mutants were isolated, and the generality of a shoot-derived signaling system has been confirmed (Kawaguchi et al. 2002; Penmetsa and Cook 1997; Schauser et al. 1998; Szczyglowski et al. 1998; Wopereis et al. 2000). The responsible genes har1 of L. japonicus and nark (nts1) of Glycine max have been determined by positional cloning, and encode receptor-like kinase containing leucine-rich repeats (LRRs), a single transmembrane domain and a serine/threonine kinase domain (Krusell et al. 2002; Nishimura et al. 2002a; Searle et al. 2003).

Studies on the autoregulation of nodulation have concentrated on mutants deficient in the shoot-derived signaling system. In addition to the shoot-derived signaling autoregulation, additional types of mutants have been recognized. The sickle mutant of Medicago is ethylene insensitive and shows hypernodulation only in the roots’ susceptible zone (Penmetsa and Cook 1997). The mutant astray of L. japonicus was reported to show hypernodulation caused by deficient nodule suppression through light signaling (Nishimura et al. 2002b). In the future, the general mechanism underlying all mutants should be interpreted and unified.

In this study, we isolated a root-determined hypernodulating mutant of L. japonicus. The mutant demonstrated almost the same phenotypic and growth characteristics as those observed in wild-type plants, indicating it was a new mutant. The novelty was further confirmed by linkage mapping and the mutant was named Ljrdh (root-determined hypernodulation) 1.


Mutations were generated by treating the seeds of L. japonicus Miyakojima MG20 in 0.4% ethyl methane sulfonate (EMS) for 8 h. Approximately 5,000 seeds were treated and M2 seeds were collected individually for each M1 plant. Approximately 120,000 seeds were sowed and cultivated on nutrient-limited soil (mixture of 14 parts Akadama soil [Kanuma Kosan, Kanuma, Japan] and one part Kreha soil [Kreha, Tokyo, Japan]). The nutrient-limited soil contained 53 mg inline image and 10 mg inline image per kg soil. The M2 plants were cultivated for 1 week and inoculated with Mesorhizobium loti MAFF 303099. Following inoculation the plants were cultivated for 3–4 weeks, and the appearance of formed nodules was observed. Plants lacking nodules, with small nodules and with an excess number of nodules were selected and cultivated for 7–10 weeks on nutrient-rich soil (Kreha soil), which contained 547 mg inline image and 58 mg inline image per kg soil. This cultivation was repeated three to five times to confirm the nodule morphology. During the screening, M3 seeds of selected M2 plants were cultivated and the appearance of nodules was followed. The selected M3 plant was crossed with a wild-type plant of Miyakojima MG20 and plants that demonstrated the same phenotype were selected for F2 generation. The mutant line was established after back-crossing two times with wild-type plants. Finally, two hypernodulation mutants OL945 and OL2168 were selected.

Linkage analysis was carried out for Nod++ mutant OL945 by crossing with L. japonicus B-129 “Gifu”. A population of approximately 600 F2 plants was used for the linkage analysis. The DNA was extracted from leaves of each F2 plant using the conventional phenol extraction method, followed by polymerase chain reaction (PCR) using SSR markers (Kazusa DNA Research Institute []) and the products were separated on 2–4% agarose gel.

For growth characterization, the plants were sowed and cultivated on nutrient-limited soil or vermiculite. In cases where vermiculite was used, nitrogen-free B&D solution (Broughton and Dilworth 1971) was supplied instead of tap water. The plants were grown in an incubator controlled for 16 h/8 h day/night cycle, at a light intensity of 150 µE s−1 m−2 at 23°C and 60% relative humidity in a Biotoron LH-200 (Nihon-Ika, Osaka, Japan). The plants were cultivated for 1 week, inoculated with M. loti MAFF 303099 and their growth was measured every 10 days. To evaluate the acetylene reducing activity, an intact plant was placed in a 20-mL glass tube with a rubber stopper and the gas phase was replaced with acetylene-containing gas (10% acetylene in air). The tube was kept at room temperature for 2 h and the amount of formed ethylene was analyzed using gas chromatography. Ten plants underwent an identical procedure and the results were used to calculate the average and standard deviation values.

For a test of nitrate sensitivity of nodulation, plants were germinated and grown in a Biotoron under the above-mentioned conditions on vermiculite supplied with nitrogen-free B&D solution in small pots for 1 week. Subsequently, the plants were inoculated with M. loti MAFF 303099. The pots were placed on chicken wire and an excess amount of B&D solution containing NH4NO3 of various concentrations (0.5–10 mmol L−1) was supplied to the pots daily. For the statistical analysis, data from 10 plants were used to calculate the average and standard deviation.

For the test of ethylene sensitivity, seedlings 1 day after germination were planted onto 0.8% agar plates containing nitrogen-free B&D solution and various concentrations of 1-amino-cyclopropane 1-carboxylic acid (ACC: 0–100 µmol L−1). The plates were kept in the dark at 23°C and after 7 days the hypocotyl length of the seedlings was measured.

For grafting, seeds of mutant OL945 and MG20 were geminated and grown in dark conditions at 23°C for 1 week on vermiculite supplied with B&D solution containing 1 mmol L−1 NH4NO3. The hypocotyl's end of shoot scion and a rootstock was taken from different plants and placed in silicone tubes with an internal diameter of 0.7–1.0 mm (Tygon, Saint-Gobain, Tokyo, Japan). The grafted plants were grown in a Biotoron under the above-mentioned conditions. Two days after grafting, M. loti MAFF 303099 was inoculated onto the plants. When describing the grafting combination, the plant name of the shoot/plant name of the root was used. For the analysis, data from five grafted plants were used to calculate the average and standard deviation.


Linkage mapping was carried for mutants OL945 and OL2168, with the responsible gene loci mapped to chromosome 1, from TM 0316 (53.7 cM) to TM 0236 (61.4 cM), for the former, and at TM 0258 of chromosome 3 for the latter. The position mapped for OL2168 was the same as that of har1 (Nishimura et al. 2002a). OL2168 developed shorter taproots and a large number of lateral roots. Considering the mapped position and the phenotypic characteristics, OL2168 was presumed to be a har1 mutant. As no mutants were reported to have loci mapped from 53.7 cM to 61.4 cM on chromosome 1, the responsible gene for mutant OL945 was considered to be a new one.

When plants were grown in nutrient-rich soil, no differences were observed between the mutant OL945 and the wild-type plant with regards to plant growth and root/shoot morphology (data not shown) During cultivation in nutrient-rich soil for 50 days, both the mutants and the wild-type plants did not form any or only a very limited number of nodules because of the high concentration of available nitrogen. In addition, growth of the mutant in nutrient-limited soil was almost identical to the wild-type plant (Fig. 1). No differences were found between OL945 and the wild-type plant with regard to the shape of the shoots, roots and leaves. The only difference was found in the number of nodulations and the growth of the plants. The mutant OL945 formed 5–10-fold more nodules than the wild-type plant (Fig. 1). As the number of nodules of OL945 increased, the acetylene reducing activity (ARA) per plant of OL945 increased, and its value was almost twice as much as the wild-type plant at 40 days post-inoculation (Fig. 1). In our study, the ARA per weight of nodules could not be measured for OL945 because some of the nodules were too small to pare off. The size of the nodules of OL945 was smaller than those of the wild-type plants, whereas the growth of OL945 was apparently more pronounced than that of the wild-type plants. A better growth rate for OL945 corresponded well with a higher ARA per plant.

Figure 1.

Growth characteristics of Lotus japonicus hypernodulation mutant OL945 (inline image) and wild-type MG20 (inline image). Data are the means of 10 plants and the error bars indicate the standard deviations. Asterisks indicate significant differences between the wild-type MG20 and OL945 using a t-test (*P < 0.01).

The nitrogen sensitivity of nodulation was compared between the mutant OL945 and wild-type plant MG20. In the B&D solution, which contained 0.5–3 mmol L−1 NH4NO3, both OL945 and the wild-type plant developed nodules, whereas at a concentration greater than 5 mmol L−1, nodules were not observed (Fig. 2). No difference was found in the suppression of nodulation between OL945 and wild-type plants.

Figure 2.

Resistance of nodulation to nitrogen concentration on vermiculite for Lotus japonicus hypernodulation mutant OL945 and wild-type MG20. Data are the means of 10 plants and the error bars indicate the standard deviations.

Ethylene sensitivity was estimated by applying its precursor ACC. Inhibition of cell elongation in the hypocotyls and root was found in both the mutant OL945 and the wild-type plant (Fig. 3).

Figure 3.

Ethylene sensitivity of Lotus japonicus hypernodulation mutant OL945 and wild-type MG20. ACC, 1-amino-cyclopropane 1-carboxylic acid.

The mutant OL945 showed nodule formation 5–10-fold more than the wild-type plant when grown in nutrient-limited soil. Three different hypernodulation mutants have been reported for L. japonicus. In Har1 and clavier mutants, hypernodulation was related to the shoot-derived signaling system (Krusell et al. 2002; Nishimura et al. 2002a; Oka-Kira and Kawaguchi 2006; Oka-Kira et al. 2005). The mutant astray forms approximately twice as many nodules as the wild-type plant, and is reported to have a deficiency in nodule suppression through light signaling (Nishimura et al. 2002b). In the known mutants of hypernodulation, several peculiar non-symbiotic characteristics were found that differed from the wild-type plant. The har1 plant develops a greater number of lateral roots than the wild-type plant (Kawaguchi et al. 2002; Nishimura et al. 2002a), whereas the mutant clavier is a dwarf with delayed flowering and a different morphology of leaf veins (Oka-Kira et al. 2005). The mutant astray exhibits characteristics that are associated with defects in light and gravity responses (Nishimura et al. 2002b). The non-symbiotic phenotypic characteristics of OL945 were almost identical to the wild-type plant, thus the mutant is apparently different from the three known hypernodulation mutants.

Grafting was carried out between MG20 and mutant OL945, with the results shown in Figs 4 and 5. When MG20 was used as the scion and OL945 as the rootstock, the root showed a hypernodulation phenotype. However, when mutant OL945 was used as the scion and MG20 as the rootstock, the root showed normal nodulation. Therefore, the hypernodulation of OL945 was regulated by the root. On the contrary, the hypernodulation of har1 and clavier is determined by the shoot genotype, as are most of the hypernodulation mutants of other legumes. The mutant OL945 was ethylene sensitive and apparently different from the sickle mutant of Medicago (Penmetsa and Cook 1997); therefore, it was designated as a novel mutant. The above-mentioned phenotypic characteristics and gene mapping strongly suggest that the mutant OL945 is a new mutant of L. japonicus. The mutant OL945 was named rdh (root-determined hypernodulation) 1.

Figure 4.

Grafting between Lotus japonicus hypernodulation mutant OL945 and wild-type MG20.

Figure 5.

Grafting between Lotus japonicus hypernodulation mutant OL945 and wild-type plant MG20. MG20/945 indicates MG20 shoots grafted onto roots of OL945. Data are the means of 10 plants and the error bars indicate the standard deviations. Data followed by different letters (a and b) are significantly different at P < 0.01.

The growth of mutant OL945 is apparently enhanced by higher ARA, which would be brought from the hypernodulation phenotype. This kind of growth enhancement was not found in the other hypernodulation mutants. Although the responsible gene for OL945 was not determined, the gene's locus would have an important role in legume breeding.

The mechanism of hypernodulation has been intensively studied in the shoot-determined hypernodulation system and is considered to be a bidirectional system of two long-distance signals: an infection signal and an autoregulation signal (Oka-Kira and Kawaguchi 2006). Nonetheless, information about this system is still scarce. Many additional studies, including studies using a root-determined system, will be required for better understanding of the autoregulation mechanism.


This study was supported by the Promotion of Basic Research Activities for Innovative Biosciences (PROBRAIN) of the Bio-oriented Technology Research Advancement Institution (Brain) of Japan. We sincerely thank Dr Masayoshi Kawaguchi for his very kind suggestions.