A shared gene drives lateral root development and root nodule symbiosis pathways in Lotus

See allHide authors and affiliations

Science  22 Nov 2019:
Vol. 366, Issue 6468, pp. 1021-1023
DOI: 10.1126/science.aax2153

Common gene yields different structures

Nodules form on legume roots to house symbiotic, nitrogen-fixing bacteria. Lateral roots, characteristic of a much broader range of plants, extend to take up nutrients and water from the soil. Soyano et al. found common ground in the developmental pathways that build nodules and lateral roots (see the Perspective by Bishopp and Bennett). Evidence from Lotus japonicus, a legume that can fix atmospheric nitrogen, shows that the nodule-forming pathway shares components with the lateral root pathway.

Science, this issue p. 1021; see also p. 953


Legumes develop root nodules in symbiosis with nitrogen-fixing rhizobial bacteria. Rhizobia evoke cell division of differentiated cortical cells into root nodule primordia for accommodating bacterial symbionts. In this study, we show that NODULE INCEPTION (NIN), a transcription factor in Lotus japonicus that is essential for initiating cortical cell divisions during nodulation, regulates the gene ASYMMETRIC LEAVES 2-LIKE 18/LATERAL ORGAN BOUNDARIES DOMAIN 16a (ASL18/LBD16a). Orthologs of ASL18/LBD16a in nonlegume plants are required for lateral root development. Coexpression of ASL18a and the CCAAT box–binding protein Nuclear Factor-Y (NF-Y) subunits, which are also directly targeted by NIN, partially suppressed the nodulation-defective phenotype of L. japonicus daphne mutants, in which cortical expression of NIN was attenuated. Our results demonstrate that ASL18a and NF-Y together regulate nodule organogenesis. Thus, a lateral root developmental pathway is incorporated downstream of NIN to drive nodule symbiosis.

Root nodule symbiosis in legumes allows host survival in nitrogen-limiting conditions and partakes in the nitrogen cycle in terrestrial ecosystems. This symbiosis has evolved through the co-option and rearrangement of signaling pathways, following predisposition in a single ancestor of the nitrogen-fixing angiosperm clade (13). It is presumed that nodulation-specific factors, such as Lotus japonicus NODULE INCEPTION (NIN), function downstream of early signaling modules (common symbiosis pathway) recruited from arbuscular mycorrhizal symbiosis, which is widely distributed in plants (Fig. 1A) (46). NIN is related to proteins involved in responses to nitrate (7). Ectopic expression of NIN and its target Nuclear Factor-Y (NF-Y) subunit genes NF-YA1 and NF-YB1 alters development of lateral root primordia and activates cortical cell division for nodule organogenesis (8), implying that NIN and its target factors link nodule development programs with lateral root developmental programs.

Fig. 1 Rhizobial infection activated ASL18a depending on NIN.

(A) Schematic of a working hypothesis. AM, arbuscular mycorrhizal. (B) Read coverage in ASL18 genes obtained by ChIP-seq analysis with NIN precipitation. Boxes indicate exons and arrowheads indicate putative NBSs (S1 and S2). bp, base pairs. (C and D) Quantitative reverse transcription polymerase chain reaction analyses of ASL18 expression. (C) Wild-type (Gifu B-129) and nin-2 roots inoculated with M. loti (n > 10 plants for each biologial replicate). dai, days after inoculation. (D) Gifu B-129 roots transformed with either an empty vector or Pro35S:NIN-GR were treated with 10 μM dexamethasone (n > 10). *P < 0.05, **P < 0.01 [one-way analysis of variance (ANOVA) with Tukey’s post hoc test] versus wild-type mock (C) and Pro35S:NIN-GR 0 hours (D). Data are mean ± SD of three biological repeats.

To identify transcription factors that influence cell division with NF-Y, we searched for genes whose transcription is induced in response to rhizobia among NIN target candidates found by a chromatin immunoprecipitation sequencing (ChIP-seq) analysis (9), and we further overexpressed them in L. japonicus roots. Only one gene, ASYMMETRIC LEAVES 2-LIKE 18/LATERAL ORGAN BOUNDARIES DOMAIN 16a (ASL18/LBD16a), stimulated cell division when co-overexpressed with NF-YA1. ASL18 genes have been duplicated at least once in an ancestral legume lineage (fig. S1). One or two NIN-binding nucleotide sequences (NBS-S1, or both NBS-S1 and NBS-S2) were found in ASL18b and ASL18a introns, respectively (Fig. 1B and fig. S2) (10). NBS-S1 and its flanking nucleotide sequences were conserved in leguminous ASL18 introns—particularly in Papilionoideae, with the exception of a few species—but were not observed in nonleguminous orthologs (fig. S1). Hence, the evolution of NBS in ASL18 intron sequences could have played a key role in recruitment of this lateral root regulator into the nodule signaling pathway in legumes. NIN was required for ASL18a expression in response to inoculation with Mesorhizobium loti (Fig. 1C). Furthermore, dexamethasone treatment of roots expressing NIN fused with a glucocorticoid receptor (NIN-GR) (8) increased ASL18a expression within 4 hours (Fig. 1D and fig. S3).

Spatial expression patterns of ASL18 genes were investigated using translational fusion with β-glucuronidase (GUS) reporter (fig. S4A; ProASL18a:ASL18a-GUS and ProASL18b:ASL18b-GUS). Both translational fusions were expressed in early lateral root primordia derived from the pericycle (Fig. 2, A and B). Lateral root densities exhibited by asl18a knockout plants were lower than those of wild-type plants (Fig. 3A and fig. S5), which was consistent with the general function of ASL18/LBD16 (1113). The ASL18a promoter was responsible for expression in lateral root primordia and for response to auxin (Fig. 2, C and D, and fig. S6). In the presence of rhizobia, ProASL18a:ASL18a-GUS was expressed at infection foci in the root epidermis and nodule primordia formed in the cortex, similar to expression patterns of NIN and NF-Y subunit genes (Fig. 2, E and F, and fig. S4) (8). ProASL18b:ASL18b-GUS showed less expression at the basal region of the nodule primordia (Fig. 2G and fig. S4). The ASL18a intron was sufficient for conferring expression in nodule primordia and its transcription induced by NIN, whereas the ASL18a promoter was also active in primordia (Fig. 2, H to J, and figs. S4 and S7). Thus, multiple pathways are connected with ASL18a transcription downstream of NIN. The number and size of asl18a mutant nodules were reduced compared with those of the wild type, when KNO3 was supplemented for partial inhibition of nodulation (Fig. 3B). This suggested that ASL18a was involved in nodule growth. The weakness of the asl18a phenotype is probably due to redundancy as observed in Arabidopsis (14). Indeed, nodule and lateral root development were inhibited when ASL18a was expressed as a fusion protein with an artificial repressor domain, SRDX, in hairy roots (fig. S8). ASL18a fused with a 35S minimal promoter partially suppressed the asl18a nodule phenotype (fig. S9).

Fig. 2 Spatial expression patterns of ASL18 genes.

GUS expression in lateral root primordia (A to D), infected root hairs (E), and nodule primordia (F to J). [(E) to (J)] Images merged with fluorescence from DsRed-labeled M. loti. Roots were transformed with ProASL18a:ASL18a-GUS [(A), (E), and (F)]; ProASL18b:ASL18b-GUS [(B) and (G)]; ProASL18a:ASL18a(cDNA)-GUS [(C) and (H)]; Pro35Sminimal:ASL18a-GUS [(D) and (I)]; and Pro35Sminimal:ASL18a(cDNA)-GUS (J) (see fig. S4A). Scale bars: 0.2 mm.

Fig. 3 ASL18a is involved in both lateral root and nodule development.

(A) Lateral root densities and primary root length (n > 15 plants) of wild type, als18a, and nf-ya1-5 nf-yb1-1 (14-day-old). (B) Numbers (n = 15 plants) and diameters (n > 37 nodules) of nodules formed under conditions supplemented with KNO3 (15 dai). (C) Nodule and primordium numbers (n > 20 plants) of multiple mutants between asl18a-1, nf-ya1-5, and nf-yb1-1. One-way ANOVA with Tukey’s post hoc test was used. Different letters represent classes with significant difference (P < 0.05). *P < 0.05, **P < 0.01. Scale bars: 1 cm for (A), 1 mm for (B).

The asl18a mutations enhanced nodulation phenotypes of nf-y subunit mutants. Nodule development was delayed and nodule number was reduced in nf-ya1 (8, 15). Nodule development was affected more severely in nf-ya1 nf-yb1 double mutants (figs. S10 and S11). We attribute enhancement of the nodulation phenotype to functional redundancies with other NF-Y subunits (16). Development of nodule primordia in asl18a nf-ya1 nf-yb1 triple mutants was delayed further, and the numbers of primordia and visible cortical division sites were approximately half of those observed in nf-y double mutants (Fig. 3C and fig. S11). This suggested the involvement of ASL18a in nodule development from early stages. Further, ASL18a seemed to genetically interact with NF-Y during nodule development. In contrast, nf-y mutations did not influence lateral root densities (Fig. 3A).

NF-Y requires other factors, including pioneer transcription factors, for an activation of its targets (17). Lotus NF-Y subunits interacted with ASL18 proteins in vitro and in planta (Fig. 4A and fig. S12). NF-Y subunits were overexpressed with or without ASL18a in roots (fig. S13). Double expression of NF-YA1 and NF-YB1 increased lateral root densities to twice those of empty vector controls (Fig. 4C and fig. S14) (8). ASL18a alone exerted no effect. Coexpression of ASL18a with both NF-Y subunits increased lateral root densities sixfold over controls, which was compatible with the protein interactions. Furthermore, roots ectopically expressing ASL18a and NF-YA1 generated bumps (fig. S14 and table S1). Likewise, triple overexpression showed bumps in both wild type and nin-9 (Fig. 4, C to E). Thus, ASL18a stimulated cell division in collaboration with NF-Y subunits. This effect was not specific to legumes (fig. S15). However, it did not increase nodule numbers (fig. S14). To examine whether ectopic cell division is associated with symbiotic events, we expressed ASL18a and NF-Y subunits in daphne mutants, in which a chromosomal translocation upstream of NIN diminishes its expression in root cortex, thereby the mutant roots host infection threads in root epidermis but do not produce nodule primordia (18). Expression of ASL18a with NF-Y subunits led to the appearance of infected nodules on daphne roots (Fig. 4, F and G, and table S1). Infection threads penetrated into nodules formed in daphne roots, and rhizobia were released into host cells when ASL18a was expressed with NF-YA1 (Fig. 4H). Although the efficiency of production of infected nodules was higher in roots coexpressing ASL18a and both NF-Y subunits, inhibition of rhizobial release suggested that a correct expression pattern is required for endosymbiosis (Fig. 4I).

Fig. 4 Interaction of ASL18a with NF-Y stimulated lateral root formation and ectopic cell division.

(A) Pulldown assay in vitro. (B) Lateral root densities of wild-type (MG-20) roots constitutively expressing ASL18a, NF-YA1, and NF-YB1 (n = 30 roots). Different letters represent classes with significant difference (P < 0.05, one-way ANOVA with Tukey’s post hoc test). (C to E) Bumps formed in MG-20 (C) and nin-9 (D) roots coexpressing ASL18a with both NF-Y subunits. (E) A longitudinal section of (C). (F to I) Empty vector–transformed daphne roots (F) or a construct to constitutively express ASL18a with either NF-Y subunits [(G) and (I)] or NF-YA1 (H) were inoculated with M. loti (4 weeks). [(H) and (I)] Sections of infected nodules stained with toluidine blue. Arrows in (G) and (I) indicate infected nodules and infection threads, respectively. Scale bars: 0.2 mm for (C) to (E), 2 mm for (F) and (G), 0.1 mm for (H) and (I).

The evolutionary origin of root nodules has been previously discussed (19, 20). Here, we show that a gene involved in lateral root development is co-opted for nodule organogenesis downstream of NIN. Replacement of NIN function by ASL18a in collaboration with NF-Y suggested the recruitment of ASL18a in organogenesis. An organogenesis-regulating molecular network has evolved through the interplay between the nodulation-specific and lateral root developmental pathways. Our findings thus clarify how legumes acquired the ability to produce root nodules.

Supplementary Materials

Materials and Methods

Figs. S1 to S15

Tables S1 to S3

References (2130)

References and Notes

Acknowledgments: We thank the National Institute for Basic Biology’s Functional Genomics Facility, Model Plant Research Facility, and Spectrography and Bioimaging Facility for technical support and A. Tokairin and A. Oda for technical assistance. Funding: This work was funded by a RIKEN Incentive Research Grant to T.S., and by KAKENHI (16K08149) and the Next Generation World-Leading Researchers grant (GS029) from JSPS to T.S. and M.H., respectively. Author contributions: T.S. and M.H. designed the study. T.S. performed experiments, with support from Y.S., M.K., and M.H., and analyzed data. T.S. wrote the manuscript. Competing interests: The authors declare no competing interests. Data and materials availability: All data are available in the main text or the supplementary materials.

Stay Connected to Science

Navigate This Article