klavier (klv), A novel hypernodulation mutant of Lotus japonicus affected in vascular tissue organization and floral induction

Authors

  • Erika Oka-Kira,

    1. Department of Biological Sciences, Graduate School of Science, The University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo 113-0033, Japan,
    2. Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency, 4-1-8 Honcho, Kawaguchi, Saitama 332-0112, Japan,
    Search for more papers by this author
    • These two authors contributed equally to this work.

  • Kumiko Tateno,

    1. Institute for Biomolecular Science, Faculty of Science, Gakushuin University, Mejiro 1-5-1, Toshima-ku, Tokyo 171-8588, Japan,
    Search for more papers by this author
    • These two authors contributed equally to this work.

  • Kin-ichiro Miura,

    1. Institute for Biomolecular Science, Faculty of Science, Gakushuin University, Mejiro 1-5-1, Toshima-ku, Tokyo 171-8588, Japan,
    Search for more papers by this author
  • Tatsuya Haga,

    1. Institute for Biomolecular Science, Faculty of Science, Gakushuin University, Mejiro 1-5-1, Toshima-ku, Tokyo 171-8588, Japan,
    Search for more papers by this author
  • Masaki Hayashi,

    1. Faculty of Horticulture, Chiba University, Matsudo 648, Matsudo, Chiba 271-8510, Japan,
    Search for more papers by this author
  • Kyuya Harada,

    1. Faculty of Horticulture, Chiba University, Matsudo 648, Matsudo, Chiba 271-8510, Japan,
    Search for more papers by this author
  • Shusei Sato,

    1. Kazusa DNA Research Institute, Kazusa-kamatari 2-6-7, Kisarazu, Chiba 292-0818, Japan,
    Search for more papers by this author
  • Satoshi Tabata,

    1. Kazusa DNA Research Institute, Kazusa-kamatari 2-6-7, Kisarazu, Chiba 292-0818, Japan,
    Search for more papers by this author
  • Naoya Shikazono,

    1. Department of Radiation Research for Environment and Resources, JAERI, Watanuki 1233, Takasaki, Gunma 370-1292, Japan, and
    Search for more papers by this author
  • Atsushi Tanaka,

    1. Department of Radiation Research for Environment and Resources, JAERI, Watanuki 1233, Takasaki, Gunma 370-1292, Japan, and
    Search for more papers by this author
  • Yuichiro Watanabe,

    1. Department of Life Sciences, Graduate School of Arts and Sciences, University of Tokyo, Komaba 3-8-1, Meguro-ku, Tokyo 153-8902, Japan
    Search for more papers by this author
  • Izumi Fukuhara,

    1. Department of Biological Sciences, Graduate School of Science, The University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo 113-0033, Japan,
    2. Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency, 4-1-8 Honcho, Kawaguchi, Saitama 332-0112, Japan,
    Search for more papers by this author
  • Toshiyuki Nagata,

    1. Department of Biological Sciences, Graduate School of Science, The University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo 113-0033, Japan,
    Search for more papers by this author
  • Masayoshi Kawaguchi

    Corresponding author
    1. Department of Biological Sciences, Graduate School of Science, The University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo 113-0033, Japan,
    2. Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency, 4-1-8 Honcho, Kawaguchi, Saitama 332-0112, Japan,
    Search for more papers by this author

(fax +81 3 5841 4458; e-mail masayosi@biol.s.u-tokyo.ac.jp).

Summary

A novel hypernodulation mutant line was isolated from Lotus japonicus Miyakojima MG-20 by irradiation with a helium ion beam. This mutant, named klavier (klv), had roots that were densely covered with small nodules. The nodulation zone of klv was significantly wider than that of the wild type. Grafting experiments showed that klv is impaired in the long-distance shoot-to-root autoregulatory mechanism. Thus the shoot genotype was found to be responsible for the negative regulation of nodule development by KLV. Nodulation of klv showed a higher tolerance to nitrogen (KNO3) than the wild type, which is a common feature of hypernodulating mutants. In addition to an increased number of nodules, the klv mutant showed convex leaf veins on the adaxial leaf surface, markedly delayed flowering and dwarf phenotypes. Microscopic examination of the leaf veins revealed that they were discontinuous. Other phenotypes such as fasciated stems, increased number of flowers and bifurcated pistils were also frequently observed in the klv mutant. Among these phenotypes, hypernodulation, aberrant leaf vein formation and significantly delayed flowering were all linked in a monogenic and recessive manner, indicating that these phenotypes are caused by either a single mutation, or tightly linked mutations. KLV was mapped within 0.29 cM on the long arm of chromosome 1.

Introduction

Nodulation in leguminous plants is a process of tissue development accompanied by bacterial invasion. The number of nodules is tightly regulated, as too many nodules consume a lot of photosynthates and consequently impair plant growth. Host plants therefore possess a negative feedback mechanism in which earlier nodulation events systemically suppress the development of younger nodules (Kosslak and Bohlool, 1984; Nutman, 1952; Pierce and Bauer, 1983). This systemic feedback regulation is also termed ‘autoregulation of nodulation’. Identification of the soybean hypernodulating mutant nts1, which forms considerably more nodules than its wild-type counterpart (Carroll et al., 1985), demonstrated that host plants possess a gene or genes negatively regulating nodule development. From grafting (Delves et al., 1986; Hamaguchi et al., 1993; Sheng and Harper, 1997) and split-root experiments (Olsson et al., 1989) with nts1 mutants, it has been proposed that nts1 lacks the shoot-derived autoregulatory signal that determines the root nodulation phenotype via long-distance signalling.

The autoregulation of nodulation is thought to consist of a bidirectional system of two long-distance signals: the infection signal and the autoregulation signal. The infection signal of early nodulation events is transferred from the root to the shoot in response to rhizobial infection. On receiving the root-derived infection signal, the autoregulation signal is produced in the shoot and then transmitted to the whole root system to prevent excessive nodulation. In pea, a root-regulated hypernodulating mutant, nod3, as well as shoot-regulated mutants, sym28 and sym29, have been identified (Duc and Messager, 1989; Sagan and Duc, 1996), indicating that SYM28 and SYM29 are responsible for production of the autoregulation signal, while NOD3 may be involved in production of a root-derived signal or the perception of the shoot-derived signal in the root. Recently, van Brussel et al. (2002) clearly showed that the systemic regulation of nodulation could be triggered by the application of mitogenic Nod factor using the Vicia split-root system, suggesting that plants sense the bacterial infection through the perception of mitogenic Nod factor.

Hypernodulating mutants have also been reported in the model legume Lotus japonicus (Kawaguchi et al., 2002; Schauser et al., 1998; Szczyglowski et al., 1998; Wopereis et al., 2000) and Medicago truncatula (Penmetsa and Cook, 1997; Penmetsa et al., 2003; Schnabel et al., 2003). Grafting experiments of hypernodulation mutants, har1 of Lotus and sunn of Medicago, have confirmed the generality of the underlying regulatory mechanism, where nodule development is controlled through long-distance signalling between root and shoot (Krusell et al., 2002; Nishimura et al., 2002a; Penmetsa et al., 2003).

The HAR1 and NARK (NTS1) genes have been cloned by positional cloning, and have been determined to encode a receptor-like kinase containing 21 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). These genes showed the highest level of similarity to the Arabidopsis LRR receptor-like kinase CLAVATA 1 (CLV1). CLV1 is expressed in the shoot apical meristem (SAM). In contrast, HAR1 and NTS1 genes are expressed in the shoots and roots, but strongly repressed in the SAM (Nishimura et al., 2002a; Searle et al., 2003).

In addition to hypernodulating mutants that are deficient in long-distance signalling, different types of mutant with an increased number of nodules have also been isolated from Lotus and Medicago, but not all of them appear to be impaired in the autoregulatory pathway. For example, the sickle mutant of Medicago, which is an ethylene-insensitive mutant, shows hypernodulation only in the susceptible zone of the root through hyperinfection (Penmetsa and Cook, 1997), unlike the autoregulation mutants that have been shown to have an extended nodulation zone. The mutant astray, isolated from Lotus, appears to be deficient in nodule suppression through light signalling (Nishimura et al., 2002b). Thus the overall number of mutated loci related to autoregulation of nodulation is still limited. The isolation and characterization of a novel hypernodulating mutant are necessary to understand the systemic features of the autoregulation mechanism.

Here we report a novel Lotus mutant with an increased number of nodules. This mutant was named klavier (klv) because its nodules lined up densely along the roots, reminiscent of the keys of a klavier or piano.

Results

Identification of a novel hypernodulating mutant by ion-beam mutagenesis

A total of 15 039 M2 plants were screened by visual inspection. One mutant line displaying hypernodulation was isolated from the M2 seedlings derived from seeds irradiated with 300 Gy helium ion (He2+). This hypernodulation mutant was named klavier (klv). In addition to its hypernodulating phenotype, klv exhibited aberrant leaf veins and significantly delayed flowering. These mutant phenotypes were stably inherited through the M3 and M4 generations. When klv was back-crossed to the parental line MG-20, both symbiotic and non-symbiotic phenotypes were wild type in the F1 generation. In the F2 generation, 165 plants showed wild-type nodulation while 53 plants were hypernodulated. The observed segregation ratio was 3.11 (x2 = 0.055) and fitted the expected value of 3:1, implying that the hypernodulation phenotype of klv is regulated in a monogenic recessive manner. Two additional mutant phenotypes of klv, convex leaf veins and late flowering, were completely co-segregating with the hypernodulation characteristic, suggesting that these three distinct phenotypes are due to either a single or tightly linked mutations.

While several mutant phenotypes were observed in plants derived from C5+-irradiated seeds or from seeds irradiated with a 400 Gy He2+ beam, these phenotypes were not recoverable in subsequent generations.

The symbiotic phenotypes of klv

Figure 1(a) shows the representative nodulation phenotypes of MG-20 and klv. In contrast to MG-20, where nodules are confined to the susceptible zone, nodules of klv densely covered the entire length of the root. The number of nodules formed on klv roots was approximately four times greater than that of MG-20, when analysed 2 weeks after inoculation with Mesorhizobium loti (Figure 1b). The difference was even greater at the later time point, 3 weeks after inoculation, when klv formed 10 times more nodules than MG-20 (Figure 1b). Three weeks after inoculation, the nodulation zone of klv was 2.4 times wider than the corresponding region of the wild-type root (Figure 1c). klv formed nodules on about 80% of the main root length, and the density of nodules within the nodulation zone was found to be 3.4 times greater than in the wild type (Figure 1c,d). In general, hypernodulating mutants tend to develop smaller nodules than wild-type plants (Wopereis et al., 2000). Nodules of klv were smaller than those of MG-20 (Figure 1e). The diameter of MG-20 nodules 3 weeks after inoculation varied from 0.1 to 2.0 mm, and 37% of the nodules were >1.0 mm. In contrast, all klv nodules were <1.0 mm in diameter (Figure 1e).

Figure 1.

Symbiotic phenotype of MG-20 and klv.
Two-week-old MG-20 and klv were inoculated with Mesorhizobium loti. (a) Nodule formation 3 weeks after inoculation (scale bar, 2 cm); (b) number of nodules and nodule primordia >0.2 mm diameter; (c) length of nodulation zone; (d) nodule density in the nodulation zone 2 and 3 weeks after inoculation (error bars, SE); (e) size of nodules 3 weeks after inoculation.

The histological structure of MG-20 and klv nodules was similar. Both consisted of a nodule cortex, a nodule endodermis, and a nodule parenchyma (inner cortex), with nodule vascular bundles and nodule central tissue (Figure 2a,c). The appearance of the central tissue, which consists of a mixture of infected and uninfected cells, did not differ either, at least at the light microscopy level (Figure 2b,d).

Figure 2.

Structure of MG-20 and klv nodules.
Sections of MG-20 (a, b) and klv (c, d) nodules. Nodule vascular bundles indicated by arrows. Scale bars, 200 μm.

The non-symbiotic phenotypes of klv

In addition to hypernodulation, klv exhibited aberrant leaf veins and significantly delayed flowering. Leaf veins were convex on the adaxial leaf surface and concave on the abaxial surface, which appears to be the opposite of MG-20 leaves (Figure 3a). Transverse sections of terminal leaflets showed that the dorso-ventral organization of klv leaves was normal; the palisade parenchyma was on the adaxial side and the spongy parenchyma on the abaxial side (Figure 3b). In MG-20 leaves a clear bundle-sheath extension was observed under the main vein, and the lower epidermal cells adjacent to the bundle-sheath extension were larger than other epidermal cells (Figure 3b). In klv leaves, however, a clear caving at the abaxial side of the leaf could be seen in the transverse sections of the main vein (Figure 3b). The bundle-sheath extension was not distinct, and the cells of the abaxial epidermis under the vascular bundle were not enlarged, or were even smaller than the other epidermal cells (Figure 3b). The convexity of the leaf vein on the adaxial leaf surface was not as significant in the main vein, but could be clearly seen in the lateral veins (Figure 3b). From the observations made with cleared leaves, it was found that the vascular differentiation in klv leaves is abnormal (Figure 3c). Leaf veins were fragmented in many places, and single cells that looked like tracheary elements were formed away from the veins (Figure 3c).

Figure 3.

Phenotype of MG-20 and klv leaves.
(a) Pictures of leaves were taken from the adaxial and abaxial sides of MG-20 and klv. Scale bar, 1 cm.
(b) Transverse sections of MG-20 and klv leaves. Upper pictures show the main vein of leaves; lower pictures show a lateral vein. A clear caving at the abaxial side of the klv main vein and an embossment on the adaxial side of the lateral vein are indicated with arrows. Scale bar, 200 μm.
(c) Leaves of MG-20 and klv were cleared and observed microscopically. Arrows show disconnected tracheary elements. Scale bar, 100 μm.

klv also exhibited extremely late flowering. Wild-type MG-20 plants start flowering about 1–1.5 months after sowing. In contrast, it took approximately 5 months for klv to flower. Importantly, all these mutant phenotypes, namely hypernodulation, aberrant leaf vein differentiation and late flowering, were manifested together in all klv hypernodulating mutant plants analysed.

Other phenotypes frequently observed in klv were fasciation of stems; increased number of flowers per peduncle; and bifurcation of pods (Figure 4). Fasciated stems were seen on about 40% of the 3-week-old klv plants (Figure 4b), but on none of the MG-20 plants of the same age. In MG-20 the number of flowers on one peduncle was one or two (Figure 4c) and the average number was 1.38 ± 0.14 (n = 13), whereas 60% of klv peduncles had more than three flowers, and up to five flowers were observed (Figure 4d). The average number of klv flowers on one peduncle was 2.96 ± 1.11 (n = 45). As far as we observed, MG-20 plants formed one pod from one flower, whereas about half of the klv flowers formed two or more pods, which is likely to be a result of the bifurcation of pistils (Figure 4e,f).

Figure 4.

Phenotypes often seen in klv mutants.
Different parts of MG-20 (a, c, e) and klv (b, d, f). Fasciation of stems was frequently seen in klv (cf. a and b). More flowers per peduncle were formed in klv compared with MG-20 (cf. c and d). Several pods in one calyx, reflecting the bifurcation of pistils, were also often seen in klv, but not in MG-20 (cf. e and f). Inset in (f) shows a bifurcated pistil. Scale bar, 1 cm.

The length of the shoot, main root, lateral roots and internodes in klv were shorter than in MG-20, even in the absence of rhizobia (Figure 5a). The hypernodulating Lotus har1 mutant is reported as having a short, highly branched root system (Kawaguchi et al., 2002; Wopereis et al., 2000). In contrast, the number of lateral roots was decreased in klv, while the number of leaves, a lateral organ of the shoot, did not differ significantly between MG-20 and klv (Figure 5b).

Figure 5.

Different growth parameters of MG-20 and klv.
Plants were grown for 3 weeks in the presence of 5 mm KNO3, without inoculation. (a) Length of shoot, main root, lateral roots, and first and second internodes; (b) number of lateral roots and leaves. Values are means of 14 (MG-20) or 11 (klv) seedlings; error bars, SE. Asterisks indicate the significance of the difference between MG-20 and klv: *, P < 0.05; **, P < 0.01.

klv mutation impairs the shoot-derived long-distance signal

In preliminary experiments we found that the success rate of grafting the Lotus ecotype MG-20 was much lower than for ecotype Gifu B-129. Only 10–15% of MG-20 grafts developed nodules, while about 50% of the grafts, using either the shoot or root of Gifu in place of the corresponding part of MG-20, nodulated. Stimulating the elongation of hypocotyls by keeping plants in darkness for 2 days also increased the success rate.

The number of nodules on klv (scion)/Gifu (root stock) was notably higher than those of the wild-type grafts (MG-20/MG-20, Gifu/Gifu, MG-20/Gifu and Gifu/MG-20; Figure 6a). Small nodules densely covered the entire root of klv/Gifu, as in klv/klv grafts or intact klv (data not shown). In contrast, the number of nodules on Gifu/klv did not differ significantly from that of any wild-type grafts (Figure 6a). The nodule number of klv/Gifu grafts was somewhat higher than in klv/klv grafts, but the difference was not significant (Figure 6a). The width of the nodulation zone was also shoot-regulated (Figure 6b). The nodulation zones of klv/klv and klv/Gifu were significantly wider than the grafts with wild-type shoots (Figure 6b). Hypernodulation of grafts with klv scions, and wild-type nodulation of klv rootstocks with wild-type scions, were observed repeatedly in three independent experiments.

Figure 6.

Grafting experiments with klv.
(a) Nodules and nodule primordia >0.2 mm diameter were counted 4 weeks after inoculation. Six grafts of each combination were made, and the number of grafts that succeeded is indicated as n.
(b) The nodulation zone was measured 4 weeks after inoculation. Each value is the mean of grafts that succeeded. Error bars, SE.

The adaxial leaf veins of klv/Gifu and klv/klv were both convex, while Gifu/klv had wild-type veins. MG-20/Gifu and MG-20/MG-20 flowered earliest, and Gifu/Gifu, Gifu/klv and Gifu/MG-20 followed. The size of the shoot as well as leaf veins and flowering time reflected the genotype of the shoot in all the graft combinations (data not shown).

Effect of nitrogen

The nodulation patterns of klv, MG-20, har1-5 (Ljsym78-2) and Gifu were analysed in the presence of increasing concentrations of KNO3. The number of nodules and nodule primordia (with diameters >0.2 mm), 3 weeks after inoculation, are shown in Figure 7. Nodulation of MG-20 was completely inhibited by KNO3 concentrations >5 mm, while klv nodulated even in the presence of 20 mm KNO3. The number of nodules formed by klv decreased gradually with increasing concentration of KNO3. However, klv was able to retain approximately 20% of the nodulation capacity, even in the presence of 20 mm KNO3. Nodulation of Gifu was not completely restricted even by a high concentration of KNO3, while under similar conditions har1 developed more nodules than klv (Figure 7).

Figure 7.

Effect of KNO3 on nodule formation.
Plants, grown for 2 weeks with B&D medium containing 5 mm KNO3, were inoculated with Mesorhizobium loti suspended in B&D solutions containing different concentrations of KNO3 (0.5–20 mm). Nodules and nodule primordia with diameters >0.2 mm were counted 3 weeks after inoculation. Each value is the mean of two to seven seedlings. Error bars, SE.

Response to ethylene

Plant hormones such as ethylene and abscisic acid are known to serve as negative regulators of nodulation (Nukui et al., 2004; Suzuki et al., 2004). Ethylene especially inhibits infection thread formation; thus ethylene insensitivity leads to an increase in nodule number through hyperinfection. In order to examine whether or not hypernodulation of klv is caused through insensitivity to ethylene, the ethylene sensitivity was analysed by applying the immediate precursor of ethylene, 1-aminocyclopropane 1-carboxylic acid (ACC). Seedlings of dicotyledonous plants grown in the dark in the presence of ethylene undergo dramatic morphological changes known collectively as the triple response: radial swelling of the hypocotyl; exaggeration in the curvature of the apical hook; and inhibition of cell elongation in the hypocotyl and root (Ecker, 1995). Both MG-20 and klv seedlings showed the triple response under treatments with 10 and 100 μm ACC (Figure 8). As the concentration of exogenous ACC increased, the growth of hypocotyls gradually became altered, suggesting that the ethylene sensitivity of klv is comparable with the wild type MG-20 (Figure 8).

Figure 8.

Effect of 1-aminocyclopropane 1-carboxylic acid (ACC) on the morphology of MG-20 and klv seedlings.
(a) Triple response to exogenous ACC (0, 1, 10 and 100 μm). Scale bar, 1 cm.
(b) Hypocotyl length under the effect of ACC. Values presented are relative to lengths without ACC. Each value is the mean of five to eight seedlings. Error bars, SE.

Fine mapping of KLV

Based on the nucleotide sequences of the TAC and BAC clones and DNA polymorphisms between two Lotus ecotypes (Miyakojima MG-20 and Gifu B-129), a number of simple sequence repeat (SSR) or derived cleaved amplified polymorphic sequence (dCAPS) markers have been generated, and localized genetically on the Lotus molecular linkage map (Asamizu et al., 2003; Kaneko et al., 2003; Kato et al., 2003; Nakamura et al., 2002; Sato et al., 2001). Using a selection of 24 SSR markers covering six Lotus chromosomes, KLV was mapped roughly to the long arm of chromosome 1 and placed between markers TM0187 and TM0220 (33.4 and 43.1 cM from the translocated region between Gifu and Miyakojima, respectively; Kawaguchi et al., 2005). 508 F2 plants and four SSR markers (TM0187, TM0410, BM1745, TM0163), located adjacent to TM0187, were used for precise mapping. We could delimit the KLV locus to a region of 0.295 cM between markers TM0410 and BM1745 (Figure 9). The genetic distance between the two markers was almost the same as the distance estimated using the MG-20–Gifu RI lines (no recombination events between the markers in 254 chromosomes), suggesting that no large chromosomal deletion or inversion occurred in the vicinity of the KLV locus.

Figure 9.

Fine mapping of KLV.

Discussion

klv is defective in the autoregulation of nodule development

klv was found to form closely spaced nodules over the entire length of the root (Figure 1). This is different from previously described mutants with increased nodule number, such as Lotus astray (Nishimura et al., 2002b), where nodules are formed within an enlarged root susceptible zone but the density of nodulation appears to be normal; and Medicago sickle, which develops an excessive number of nodules restricted to the root susceptibility zone (Penmetsa and Cook, 1997). On the other hand, the hypernodulation phenotype of klv appears to be similar to the Lotus autoregulatory mutant har1, but both mutants differ significantly with respect to their non-symbiotic root and shoot phenotypes (Wopereis et al., 2000). For example, the number of lateral roots is increased in har1, while it is decreased in klv. Also, the various shoot phenotypes of klv have not been observed in har1. Different map positions for klv (localized on chromosome 1) and har1 (localized on chromosome 3) are predicted to specify distinct functions which may be responsible for the phenotypic differences observed between these mutants.

The nodules of klv were smaller than the wild type, probably because the nutrition provided from the plant to each nodule was limited. Except for their size, klv nodules appeared to have the normal wild-type structure, suggesting that there is no defect in the development of nodules. klv shoots grafted onto wild-type roots caused hypernodulation, suggesting that the hypernodulation property of klv is regulated by the shoot (Figure 6). In other words, KLV function in the shoot is required for repression of nodule development in the root. klv shoots not only caused an increase in the total number of nodules, but also affected the width of the nodulation zone. The number of nodules was the highest in klv/Gifu grafts, not in the klv/klv combinations, probably reflecting better growth of klv/Gifu grafts. Unlike the symbiotic phenotypes, non-symbiotic phenotypes of the shoot were not affected by grafting.

A common characteristic of hypernodulation mutants reported so far is their ability to nodulate in the presence of increased KNO3 concentration (N tolerance). The number of klv nodules decreased in the presence of high N concentrations, but nodulation was not as strongly affected as in wild-type MG-20 (Figure 7). This result suggested that nodulation of klv is at least partially insensitive to N inhibition. However, har1 was found to be more tolerant than klv to externally supplied N. Whether the difference in N tolerance between klv (ecotype MG-20) and har1 (ecotype Gifu) results from different functions of the corresponding KLV and HAR1 proteins, or is the result of different genetic backgrounds, remains an open question. Isolating har1 from the MG-20 background or klv from the Gifu background may be important in order to fully resolve this question.

Studies using the soybean hypernodulation mutant nts382 have shown that both autoregulation and nitrate tolerance are regulated by the shoot phenotype (Day et al., 1989). Considering that the klv shoots were responsible for the hypernodulation phenotype of the roots, the KLV-mediated autoregulatory mechanism may also be involved in regulation of N signalling.

klv affects organ development in both root and shoot

The most distinct and stable non-symbiotic phenotype of klv was the formation of aberrant leaf veins (Figure 3). Leaf veins consist of vascular bundles, tissues that are likely to contribute to the delivery of long-distance signals. A defect in the formation of vascular tissue may have an adverse impact on proper signalling through the autoregulatory pathway. In fact, disconnection of the vascular system was observed in klv leaves (Figure 3c). When leaves from 3-week-old plants were compared, the bundle sheath extension and lower epidermal cells adjacent to the vascular tissue were abnormal in klv (Figure 3b). The defect in the vascular tissue may also cause dwarfing through the limitation of nutrient transport from the root to the shoot. Comparing the development of the vascular tissue between MG-20 and klv in more detail will be needed to test these various possibilities.

Fasciation of stems and an increase in organ number are phenotypes often observed in meristem regulation mutants, such as Arabidopsis CLAVATA mutants. The two homologous genes CLV1 and HAR1 are both involved in organ differentiation, but CLV1 regulates the SAM through short-distance signalling within the SAM (Clark et al., 1997), while HAR1 regulates the nodule and root meristems through long-distance signalling from shoot to root (Krusell et al., 2002; Nishimura et al., 2002a; Wopereis et al., 2000). KLV probably represents a novel Lotus regulator of SAM and nodule meristems.

It has been reported that the hypernodulation mutant nod4 of pea also has convex leaf veins and fasciated stems (Sidorova and Shumnyi, 2003). The linkage between aberrant leaf vein formation, fasciation and hypernodulation may be conserved in legumes. Further studies using klv may shed light on the meristem regulation of Lotus, and may connect the regulation of the SAM, leaf venation, and root and nodule meristem differentiation.

klv exhibits a late-flowering phenotype although it was isolated from the earliest flowering accession of Lotus, Miyakojima MG-20 (Kawaguchi, 2000). Despite the dwarf shoot, presumably due to inhibition of internode elongation, the number of leaves in klv appears to be normal (Figure 5). Flowering and nodulation represent two distinct developmental processes that are regulated via long-distance signalling. In the case of flowering, a hypothetical systemic signal called florigen is presumed to be generated in the leaf and transmitted to the SAM, where it mediates a developmental shift from vegetative to reproductive growth (Chailakhyan, 1937; Knott, 1934; for review see Colasanti and Sundaresan, 2000). There are at least three possibilities to explain the delay of flowering in the klv mutant. The first possibility is the impairment in vascular tissue differentiation (for example phloem differentiation) required for long-distance signalling. In this case, klv may fail to effectively transmit the systemic signal responsible for both floral induction and autoregulation of nodulation. Alternatively, KLV may be expressed in the SAM and serve as a positive regulator of flowering and the maintenance of the shoot, floral, and nodule meristems. Finally, it is possible that KLV encodes a shoot-derived signal for nodule differentiation that is also involved in flower initiation. The identification and functional characterization of KLV will be required to clarify the intriguing relationship between flowering and nodulation.

Ion-beam mutagenesis identified a novel hypernodulation mutant

The use of an ion beam for mutagenesis of Arabidopsis demonstrated that C5+ and He2+ irradiation has a higher linear energy transfer than X- and gamma-rays. Ion beam irradiation induces deletions, insertions, base-pair changes and inversions (Shikazono et al., 2005). The present study represents an attempt to use ion beam irradiation for the mutagenesis of a leguminous plant. A single symbiotic mutant, klv, was isolated from seeds irradiated with 300 Gy He2+ ion beam. Given the highly pleiotropic shoot phenotypes of klv and hypernodulation, a large chromosomal deletion was anticipated at first. However, KLV could be positioned in a narrow region within 0.295 cM by fine mapping, and the genetic distance was almost the same as that obtained from recombinant inbred (RI) lines (Gifu B-129 × Miyakojima MG-20), indicating that no large chromosomal deletion or inversion had occurred in the vicinity of the KLV locus. Assembly of a contig of transformation-competent artificial chromosome (TAC) including the KLV locus, and searching for molecular lesions in the contig, are in progress.

In general C5+, with linear energy transfer (LET) higher than He2+, is thought to be more effective in causing DNA damage. Several mutant phenotypes were also induced by C5+, but none of those phenotypes were recovered. In this case, C5+ ions with high LET and dose may have been too strong to induce symbiotic mutation. Also, the range of the C5+ ion is about 1.1 mm, while the range of the He2+ ion is about 1.7 mm. He2+ irradiation may have been more effective than C5+ because the diameter of a Lotus MG-20 seed is around 1 mm. Other accelerated ions that possess different LET and can pass completely through the Lotus seeds may be more suitable for future Lotus mutagenesis.

Experimental procedures

Plant material and screening of symbiotic mutants by ion beam irradiation

An early-flowering ecotype of Lotus Miyakojima MG-20 was used for ion-beam mutagenesis. Dry seeds of MG-20 were sandwiched between Kapton films (thickness, 7.5 μm; size, 5 cm2; Toray-Dupont, Tokyo, Japan) to make a monolayer of the seeds in order to achieve homogeneous irradiation (Tanaka et al., 1997). 50-MeV helium ion (He2+) and 220-MeV carbon ion (C5+) irradiation was conducted in an AVF cyclotron (JAERI, Takasaki, Japan). The irradiation dose was calculated from LET and the particle fluence of ions. The mean LET for He2+ and C5+ was 19.4 and 158.5 keV μm−1, respectively (taking seed thickness as 1.0 mm). The irradiation dose was 50, 100, 150, 200, 250 and 300 Gy for the C5+ beam, and 200, 300 and 400 Gy for the He2+ beam. M1 seeds were grown in an air-conditioned glasshouse at the National Institute of Agrobiological Sciences, Tsukuba, Japan. M2 seeds were collected and germinated in 10 plastic trays (35 cm long, 50 cm wide, 7 cm high) containing vermiculite, watered with B&D N-deficient medium (Broughton and Dilworth, 1971). The seedlings were then inoculated with Mesorhizobium loti MAFF30-3099 or TONO strains, about 1 l per tray, at a cell density of 107 cells ml−1. Putative mutants were screened by visual assay 1 month after inoculation. The klv mutant was isolated from the M2 generation of ion beam-irradiated seeds, as a root mutant with an increased number of nodules.

Plant growth conditions

Seeds of klv and MG-20 were scarified using sandpaper, surface-sterilized with 50% sodium hypochloride solution containing 0.02% Tween-20 for 10 min, and rinsed extensively with sterilized water. After overnight incubation in sterilized water with constant shaking, seedlings were transferred onto sterilized vermiculite in plastic pots with a cover (6 cm long and wide, 9.5 cm high). The vermiculite was supplemented with B&D medium containing 5 mm KNO3, and the plants were grown aseptically. Plants were placed under a 16-h day/8-h night cycle at a light intensity of 150 μE sec−1 m−2 at 22°C in a Biotron LH-300 (Nihon-ika Co. Ltd, Osaka, Japan).

Bacterial strains and culture conditions

Mesorhizobium loti MAFF 30-3099 was used for inoculation of rhizobia. Mesorhizobium loti was grown in YEM at 28°C for 2 days. The bacteria were washed twice with sterile water and suspended in B&D medium containing 0.5 mm KNO3 unless otherwise indicated. Plants were inoculated by flooding the roots with the bacterial suspension.

Measurement of nodulation zone

The distance between the uppermost and lowermost nodules formed on the main root was measured, and was designated the nodulation zone. The ratio of the nodulation zone was calculated by dividing its length by the length of the main root.

Histological observations

Nodulated roots or leaves were fixed and cleared overnight by incubation in a mixture of acetic acid and ethanol (1:4 v/v). The fixed samples were rinsed with 80% (v/v) ethanol, and passed through ethanol series (70, 50, 30 and 10%, v/v) for 20 min each. Samples were then left in distilled water for a further 20 min. All samples were cleared in a solution containing 160 g chloral hydrate and 25.2 g glycerol in 40 ml distilled water. The cleared samples were observed with Nomarski interference optics (BX-51; Olympus, Tokyo, Japan) and photographed using a digital camera (DP50; Olympus).

Technovit sections were prepared as follows. Nodulated roots or leaves were fixed in a mixture of acetic acid and ethanol (1:4 v/v) and embedded in Technovit 7100 embedding solution (Kulzer & Co. GmbH, Wehrheim, Germany). Sections 3 or 5 μm thick were generated and stained with 0.1% (w/v) toluidine blue dissolved in 0.1 m sodium phosphate buffer (pH 7.0), and mounted in Entellan neu (Merck Darmstadt, Germany). The sections were observed under a light microscope (BX-51; Olympus) and photographed.

Crossing MG-20 and klv

MG-20 and klv were reciprocally crossed as described by Handberg and Stougaard (1992).

Grafting experiments

Seeds of klv, MG-20 and Gifu B-129 were germinated and grown as described above. When the cotyledons opened, pots with plants were moved into darkness to enhance hypocotyl elongation. The plants were returned to normal light conditions after 48 h incubation in the dark.

Plants were taken out of the pots, dipped in B&D medium containing 0.5 mm KNO3, and cut at the middle of the hypocotyls. The hypocotyl ends of a shoot scion and a rootstock taken from different plants were placed in polyethylene plastic or silicone tubes (0.7–1.0 mm internal diameter; Tygon, Saint-Gobain K.K., Tokyo, Japan) so that their cut surfaces were in contact. Grafted plants were planted back into the pots and inoculated with M. loti 2 days after grafting. Combinations of different shoot and roots are designated as scion/rootstock. For example, klv/Gifu indicates klv shoots grafted onto Gifu roots; and MG-20/MG-20 indicates the combination of shoots and roots taken from different MG-20 plants.

Nitrate-sensitivity analysis

Two-week-old seedlings were inoculated with M. loti suspended in B&D medium containing varying concentrations of KNO3 (0.5–20 mm), and grown for an additional 3 weeks. The B&D solution was exchanged 7 and 14 days after inoculation for newly made B&D containing the same KNO3 concentration as at the original inoculation, to eliminate any change in N concentration through uptake by the plants.

Ethylene-sensitivity analysis

For ethylene-sensitivity analysis, seedlings 1 day after germination were planted onto 0.8% agar plates containing N-free B&D medium and various concentrations of ACC (0–100 μm). Plates were kept in continuous dark conditions at 25°C. After 7 days’ incubation, seedlings were scanned and the hypocotyl length was measured.

Fine mapping

The hypernodulating mutant klv was crossed with Lotus Gifu B-129. The F2 seeds were sown in vermiculite in B&D N-free nutrient solution containing M. loti MAFF 30-3099. Total DNA was extracted from leaves of F2 seedlings. The map position of klv was evaluated roughly using the following 24 SSR markers: TM0027, TM0036, TM0050, TM0017, TM0001, TM0098, TM0029 for chromosome 1; TM0056, TM0020, TM0021 for chromosome 2; TM0080, TM0005, TM0049 for chromosome 3; TM0087, TM0030, TM0046, TM0097 for chromosome 4; TM0052, TM0048, TM0034 for chromosome 5; and TM0014, TM0013, TM0045, TM0055 for chromosome 6. PCR was carried out in a GeneAmp PCR System 9700 (Applied Biosystems, Foster City, CA, USA) under the following conditions: 94°C for 5 min, then 25 cycles of 94°C for 30 sec, 54°C for 30 sec and 72°C for 30 sec. The amplified PCR products were resolved on a non-denaturing 15% polyacrylamide gel and stained with Vistra Green (Amersham Biosciences, Piscataway, NJ, USA). For precise mapping of klv, SSR markers located between TM0050 and TM0017 [TM0187, TM0410, BM1745 (forward primer: CAAAATACATACATCCCGAG; reverse primer: TTGTACACCCAATAAGTGGA), TM0163, TM0220] were used. The genetic map position of klv was calculated using the program mapmaker/exp ver. 3.0b (Lincoln et al., 1992).

Acknowledgements

We would like to thank Emiko Kobayashi, Hiroshi Kouchi and Shoichiro Akao (National Institute of Agrobiological Sciences, Tsukuba, Japan) for performing plant cultivation and genetic crosses. We also thank Tomoko Takizawa (Niigata University, Niigata, Japan) and Asuka Kuwabara (The University of Tokyo, Hongo, Japan) for technical support. We are grateful to Krzysztof Szczyglowski (Agriculture and Agri-Food Canada, London, Canada) for providing valuable comments on the manuscript prior to its submission.

Ancillary