Two members of a Nodule‐specific Cysteine‐Rich (NCR) peptide gene cluster are required for differentiation of rhizobia in Medicago truncatula nodules

SUMMARY

Legumes have evolved a nitrogen‐fixing symbiotic interaction with rhizobia, and this association helps them to cope with the limited nitrogen conditions in soil. The compatible interaction between the host plant and rhizobia leads to the formation of root nodules, wherein internalization and transition of rhizobia into their symbiotic form, termed bacteroids, occur. Rhizobia in the nodules of the Inverted Repeat‐Lacking Clade legumes, including Medicago truncatula, undergo terminal differentiation, resulting in elongated and endoreduplicated bacteroids. This transition of endocytosed rhizobia is mediated by a large gene family of host‐produced nodule‐specific cysteine‐rich (NCR) peptides in M. truncatula. Few NCRs have been recently found to be essential for complete differentiation and persistence of bacteroids. Here, we show that a M. truncatula symbiotic mutant FN9285, defective in the complete transition of rhizobia, is deficient in a cluster of NCR genes. More specifically, we show that the loss of the duplicated genes NCR086 and NCR314 in the A17 genotype, found in a single copy in Medicago littoralis R108, is responsible for the ineffective symbiotic phenotype of FN9285. The NCR086 and NCR314 gene pair encodes the same mature peptide but their transcriptional activity varies considerably. Nevertheless, both genes can restore the effective symbiosis in FN9285 indicating that their complementation ability does not depend on the strength of their expression activity. The identification of the NCR086/NCR314 peptide, essential for complete bacteroid differentiation, has extended the list of peptides, from a gene family of several hundred members, that are essential for effective nitrogen‐fixing symbiosis in M. truncatula.

Significance Statement

Our results demonstrate the requirement of NCR086/NCR314, which are members of the nodule‐specific cysteine‐rich peptide family, for effective symbiotic interaction between Medicago truncatula and the tested rhizobia strains. The NCR086/NCR314 peptides are encoded by a highly similar gene pair in M. truncatula A17 that was duplicated following the separation of lines M. truncatula A17 and Medicago littoralis R108.

INTRODUCTION

Legumes can form symbiotic relationships with nitrogen‐fixing soil bacteria, collectively termed rhizobia. The site of the nitrogen‐fixing symbiosis is a newly formed organ, the root nodule, in which the symbiotic cells are colonized by thousands of rhizobia. Bacteria are encompassed by a plant‐derived membrane forming a new subcellular organelle, the symbiosome. Intracellular bacteria multiply and undergo morphological and metabolic transitions to develop into the bacteroids that convert the atmospheric nitrogen into ammonia (Jones et al., 2007; Yang et al., 2022).

Depending on the persistent or the transient activity of root nodule meristem, indeterminate and determinate nodules are distinguished (Hirsch, 1992). Indeterminate nodules, generally formed on the roots of temperate legumes such as Medicago species, are cylindrical and composed of a developmental gradient of cells defining a characteristic distinct zonation of indeterminate nodules (Gavrin et al., 2014; Sprent & James, 2007; Vasse et al., 1990). The apical meristem (Zone I) produces cells for the subsequent zones in a proximal direction. The mitotic activity of cells ceases in the subsequent zone (infection zone; ZII) and rod‐shaped bacteria are released from infection threads to colonize the cytoplasm of the host cells. Subsequently both rhizobia and infected cells undergo morphological and metabolomic transition. The differentiation of rhizobia and symbiotic cells is completed in the interzone (IZ) and the bacteria that switched to an endosymbiotic lifestyle and began reducing atmospheric nitrogen in the nitrogen fixation zone (ZIII) of the indeterminate nodules are known as bacteroids. Rhizobia in Medicago truncatula nodules duplicate their genomes multiple times without cytokinesis, producing elongated bacteria (Mergaert et al., 2006). Simultaneously, the infected nodule cells undergo polyploidization, causing the host cells to enlarge (Vinardell et al., 2003) although this enlargement is hardly detectable in some cases (Crespo‐Rivas et al., 2016). In the nodules of legumes belonging to the inverted repeat‐lacking clade (IRLC), which contain a deletion of a 25 kb inverted repeat in their chloroplast genomes, or in Aeschynomene species, the bacteroid transition is irreversible (Czernic et al., 2015; Mergaert et al., 2006). This terminal differentiation of rhizobia is controlled by the host plant and proposed to be mediated by defensin‐like nodule‐specific cysteine‐rich (NCR) peptides (van de Velde et al., 2010). The genomes of IRLC legumes contain different sized NCR gene families, and it was shown that the morphotype of the rhizobia, the degree of their elongation, correlates with the number of NCRs found in each species. Unique among them is M. truncatula genome encoding more than 700 NCR peptides (Montiel et al., 2017; Satgé et al., 2016). NCRs are expressed almost exclusively in infected nodule cells (Guefrachi et al., 2014) and encode for peptides with N‐terminal signal peptides that target them to the symbiosome (Wang et al., 2010). In most cases, mature peptides contain four or six cysteines in conserved positions, which are essential for the symbiotic activity of certain NCRs (Horváth et al., 2015, 2023), but apart from these residues, NCR peptides have a considerable degree of diversity in their amino acid compositions and sequences, resulting in a diverse spectrum of cationic, neutral, and anionic peptides (Mergaert et al., 2003; Montiel et al., 2017).

Despite the intensive research of the abundant gene family of NCR peptides in M. truncatula, few peptides, NCR169, NCR211, NCR247, NCR343, and NCR‐new35 have been shown to be essential for bacteroid differentiation and persistence in symbiotic nodules (Gao et al., 2023; Horváth et al., 2015, 2023; Kim et al., 2015; Sankari et al., 2022; Zhang et al., 2023). All but NCR247 of these peptides were identified by genetic analysis of M. truncatula deletion mutants, which show defects in colonization of ZIII in their round‐shaped white ineffective nodules and incomplete differentiation of bacteroids. The gene‐edited knockdown allele of NCR247 also develops small and white ineffective nodules indicating the requirement of NCR247 for symbiotic nitrogen fixation in M. truncatula (Sankari et al., 2022). The studies of the allele specific interaction of M. truncatula lines and Sinorhizobium meliloti strain Rm41 identified NCR peptides NFS1 and NFS2 and extended the paradigm of NCR function in bacteroid differentiation demonstrating that certain NCRs can regulate the outcome of symbiotic nitrogen fixation in a strain‐specific manner by exploiting the antibacterial properties of NCRs (Wang et al., 2017; Yang et al., 2017).

In this study, we report the identification of the NCR086 and NCR314 peptide genes required for effective symbiotic nitrogen fixing interactions between M. truncatula and the tested rhizobia strains. The loss of NCR086 and NCR314 results in incomplete differentiation of bacteroids and blocks the invasion of the nitrogen fixation zone of the symbiotic nodule. NCR086 and NCR314, which are members of a large NCR cluster in the M. truncatula genome, encode identical mature peptides and either one can restore the symbiotic phenotype of FN9285. Our study suggests that the peptide NCR086, or its recently duplicated counterpart NCR314, is essential for the terminal transition of rhizobia into bacteroids in M. truncatula Jemalong/A17 nodules.

RESULTS

The Medicago truncatula FN9285 mutant is defective in symbiotic nitrogen fixation and mutant nodules are impaired in colonization of nitrogen fixation zone

The NF‐FN9285 (FN9285) mutant line, exhibiting an ineffective (Fix‐) symbiotic phenotype, was identified in a mutant screen of fast neutron‐bombarded population of M. truncatula A17 (Xi et al., 2013). To assay the strain dependence of the symbiotic phenotype of FN9285, mutant plants were inoculated with four different rhizobia strains. The gross and the nodulation phenotypes of the wild‐type and FN9285 mutant plants were analyzed under symbiotic conditions 3 weeks postinoculation (wpi) with S. meliloti strains 1021 and 2011 (Sm2011), Sinorhizobium medicae strains ABS7 and WSM419 (Figure 1a–d). Mutant plants exhibited the symptoms of nitrogen deficiency, such as chlorotic leaves, stunted shoots, with all tested rhizobia strains, including the most compatible S. medicae WSM419 strain (Terpolilli et al., 2008). At 3 wpi, slightly elongated and white nodules were formed on the roots of FN9285 plants, indicating the lack of leghaemoglobin (lower panels of Figure 1a–d). In contrast, wild‐type plants developed cylindrical pink nodules on their roots with all four rhizobia strains. The nitrogen deficiency of FN9285 plants has certainly increased the number of nodules 5 wpi on mutant roots (Figure 1e) and apparently reduced the dry weight of the aerial part of mutant plants compared with wild‐type plants (Figure 1f). The acetylene reduction assay (ARA) was used as a measure of nitrogenase activity in wild‐type and mutant nodules 5 wpi with the four rhizobia strains. The analysis revealed the complete loss of nitrogenase activity in FN9285 nodules, but a substantial amount of ethylene production was detected in all wild‐type plants (Figure 1g). These findings indicate that mutant FN9285 is not able to establish effective symbiotic with none of the four Sinorhizobium strains.

Figure 1.

The symbiotic nitrogen fixation is defective in Medicago truncatula mutant FN9285 analyzed at 3 (gross habit and nodule morphology) and 5 (nodule number, shoot dry mass, nitrogenase activity) weeks post inoculation.

FN9285 inoculated with Sinorhizobium meliloti strains 1021 (a) and 2011 (b), Sinorhizobium medicae strains ABS7 (c) and WSM419 (d), shows the symptoms of nitrogen starvation (retarded growth, chlorotic leaves) and develops slightly elongated white nodules. In contrast, wild‐type (WT) plants displayed vibrant green shoots and developed cylindrical pink nodules.

(e) A significant increase in the nodule number on the roots of NF9285 mutant, irrespective of the rhizobia strain used for inoculation.

(f) The dry weight of the aerial part of FN9285 mutant plants was apparently reduced compared to wild‐type plants.

(g) The activity of bacterial nitrogenase was measured by acetylene reduction assay (ARA). Wild‐type nodules exhibited a substantial amount of ethylene production, whereas FN9285 nodules showed a complete loss of nitrogenase activity with all tested rhizobia strains.

The data represent the mean + SEM of three biological replicates of the measurements of nodule numbers and dry weight with all four rhizobia strains using minimum 30 plants. ARA was performed in four biological and three technical replicates. **P ≤ 0.01, ***P ≤ 0.001 using the student's test. Bars: 2 cm for plants and 200 μm for nodule images.

To investigate the colonization of symbiotic nodules, mutant and wild‐type plants were inoculated with a Sm2011 derivative that carries plasmid pXLDG4 and thus constitutively expresses the lacZ bacterial marker gene. Longitudinal nodule sections were stained for β‐galactosidase activity 14 and 21 dpi, and the presence of rhizobia was analyzed by light microscopy. At 21 dpi, the wild‐type nodules showed the characteristic zonation of indeterminate nodules of M. truncatula and cells in zones II, IZ, and III were colonized by rhizobia (Figure 2a). The zones in the less developed mutant nodules were less apparent at 14 dpi but they could be clearly demarked at 21 dpi. Histochemical staining for lacZ activity of mutant nodules detected rhizobia‐infected cells in the infection zone and interzone of mutant nodules. The region corresponding to the nitrogen fixation zone was not completely devoid of rhizobia, but X‐Gal‐stained bacteria retained in infection threads and located intracellularly (Figure 2a), a very similar habitus identified in the saprophytic zones of alfalfa nodules (Timmers et al., 2000), were found in FN9285 nodules. This diffuse β‐galactosidase staining was more prominent in mutant nodules at 21 dpi and fully colonized cells were found only in few cell layers of the infection and differentiation zones indicating that the defective gene in FN9285 is required for the complete differentiation and the persistence of rhizobia in FN9285 nodules.

Figure 2.

The deficiency in Medicago truncatula mutant FN9285 blocks the symbiotic interaction with Sinorhizobium meliloti 2011 (Sm2011) expressing the lacZ marker gene.

(a) Longitudinal sections of wild‐type and mutant nodules stained for β‐galactosidase activity at 14 and 21 days postinoculation (dpi) show the normal colonization of nodule zones characteristic for WT indeterminate nodules. In contrast, only the infection and interzone were colonized by rhizobia in FN9285 nodules. The inset shows a higher magnification of the 2 wpi‐mutant nodule. The blue staining in the region of mutant nodules corresponding to the ZIII of WT nodules shows bacteria retained in infection threads and saprophytic rhizobia released from infection threads.

(b) Laser confocal microscopic analyses of colonization of symbiotic cells and bacteroid morphology on SYTO13‐stained longitudinal nodule sections at 14 dpi with Sm2011. The lack of colonized cells in ZIII of FN9285 nodules indicates a defect in the persistence of rhizobia. Higher magnification revealed disordered rhizobia starting from the IZ cells and bacteria retained in infection threads were detected in ZIII of mutant nodules.

(c) Bacteroid size distribution isolated from wild‐type and mutant nodules at 17 dpi, stained with propidium iodide (PI) and imaged using confocal laser scanning microscopy. The bacteroids from FN9285 mutant nodules exhibited a higher percentage of shorter bacterial cells compared with bacterial cells from Mtsym20, FN9363, and wild‐type nodules. Measurements were conducted on at least 1000 bacterial cells, and their distribution across length classes is presented in a line chart. The values on X‐axis indicate bacterial length (μm), and the Y‐axis represents the corresponding percentage distribution.

(d) The relative size and DNA content of Sinorhizobium meliloti 2011 free living cells, bacteroids isolated from wild‐type (WT) and mutant nodules 17 days postinoculation (dpi) were measured by flow cytometry. Bacteroid size was assessed by the forward light scatter (FSC), while endoreduplication was determined by fluorescence intensity. The Sm2011 bacteroid populations from FN9285 mutant nodules exhibited a shift towards smaller size and lower ploidy levels compared to wild‐type and other ncr mutant nodules of Mtsym20 and FN9363.

Bars: 200 μm for whole nodules; 200 μm for the inset; 20 μm for enlarged images. b, bacteroid; it, infection thread; IZ, interzone; n, nucleus; v, vacuole; ZI, meristem; ZII, infection zone; ZIII, nitrogen fixation zone.

Nodules store amyloplasts as carbon sources for rhizobia and starch granules are highly accumulated in the interzone of Medicago nodules and to a decreasing extent in the other nodule region proximal to the root (Hostak et al., 1987). Starch distribution is referred as a marker for changes in cellular metabolism during nodule differentiation (van Schadewijk et al., 2020). Sections of wild‐type and mutant nodules were stained with Lugol's solution to visualize starch granules. We observed starch accumulation throughout the infected tissues of wild‐type nodules but cells in the region corresponding to the nitrogen fixation zone of mutant nodules were without starch (Figure S1a,b), indicating the block of metabolic transition of FN9285 nodules.

To investigate nodule colonization and bacteroid morphology more thoroughly, mutant and wild‐type nodule sections were stained with the fluorescent nucleic acid‐binding dye SYTO13 at 21 dpi with Sm2011 and examined using laser‐scanning confocal microscopy. Nodule sections of wild‐type nodules showed rhizobia in all infected tissues but bacteria were observed only in the infection and interzone of mutant nodules, and dispersed fluorescent signals were detected in the nitrogen fixation zone (Figure 2b), similarly to what we detected with X‐Gal staining (Figure 2a). At higher magnification, SYTO13‐stained nodule sections revealed that infected cells in ZII had a similar shape in mutant and wild‐type nodules. In ZII cells, bacterial release was normal and in the cells of ZII, slightly elongated bacteria were observed indicating the initiation of bacteroid development. However, rhizobia were shorter and observed in a disorderly distribution in mutant IZ cells compared with the elongated bacteroids positioned towards the central vacuoles of infected wild‐type cells. In the nitrogen fixation zone of wild‐type nodules, symbiotic cells with large vacuoles, differentiated bacteroids, and bright fluorescent bacteria in infection threads were detected (Figure S1c). In contrast, we found that no vacuoles were visible in the cells in the distal part of ZIII of mutant nodules. These cells were full of disorganized and undifferentiated rod‐shaped bacteria released from the infection threads. The mutant cells in ZIII proximal to the root were devoid of bacteroids, and only infection threads with retained bacteria were detected on the images. From these analyses, we concluded that the gene defective in FN9285 is required for complete differentiation and persistence of rhizobia in nodule cells.

Bacteroid differentiation is incomplete in FN9285 mutant nodules

The fully differentiated bacteroids are multiploids and elongated in the nitrogen‐fixing cells of the indeterminate nodules of M. truncatula (Mergaert et al., 2006). The images of SYTO13‐stained mutant IZ and ZIIId cells showed less elongated rhizobia compared with wild‐type nodules (Figure 2b; Figure S1c). To assess the extent of bacteroid differentiation, the length and the ploidy level of bacteroids were measured. Bacteria were isolated from FN9285 and other ineffective symbiotic mutants, Mtsym20 (defective in the nodule‐specific cysteine‐rich gene NCR‐new35), FN9363 (deficient in NCR343) (Horváth et al., 2023), all of them impaired in supporting complete bacteroid differentiation, and wild‐type nodules 16 dpi with Sm2011, stained with propidium iodide (PI) and imaged by confocal laser scanning microscopy. Size of at least 1000 bacteroids was measured on the images using the ImageJ software and the percentage of each size category of bacteria was plotted (Figure 2c). The bacteroid population isolated from the wild‐type nodules had a higher proportion of longer cell sizes compared with the mutant samples, and we also found that the size of bacteroids isolated from FN9285 nodules was shifted towards shorter sizes compared with rhizobia isolated from the two previously isolated mutant nodules defective in NCR343 and NCR‐new35.

The size and DNA content of the bacteroids were further analyzed by flow cytometry. The diagram showing forward scattering (FSC), which is considered to measure the relative cell size, shows that free living rhizobia were small, while bacteroids of the three mutants developed longer, but they did not reach the length of bacteroids isolated from wild‐type nodules (Figure 2d). The bacteroids isolated from FN9285 were slightly smaller in size compared to the bacteroids isolated from the other two mutants, similarly to the result identified by length measurements, which might indicate that the elongation of rhizobia in FN9285 nodules is arrested at a slightly earlier stage of differentiation compared with the two other mutants. We also analyzed the fluorescence intensity of the PI‐stained bacteria, which is proportional to the genome amplification of the bacterial cells and can be used as a measure of bacteroid differentiation. The peak of fluorescence intensity of PI‐stained bacterial cells of free living Sm2011 bacteria was the least and the bacteroids from wild‐type nodules showed the highest ploidy level (Figure 2d). In consistent with the length measurements, bacteria isolated from the mutant nodules showed intermediate fluorescence intensity, indicating that the genome amplification was incomplete in mutant nodules. From these results, we concluded that the bacteroid differentiation in FN9285 nodules is blocked before completing the mature stage.

FN9285 is deficient in a cluster of NCR genes

A combined strategy of genetic mapping and transcriptome analysis of the symbiotic mutant FN9285 was applied to identify the defective gene or genes. FN9285 plants were crossed to M. truncatula genotype A20, and the genetic mapping carried out on the F2‐segregating populations identified the symbiotic locus of FN9285 on chromosome 3 (LG 3) between the genetic markers Mtb319 and Chr3_6,7M (Figure S2a). To identify the gene deficient in FN9285, we performed RNA‐seq experiments using 14‐day‐old nodules of mutant FN9285 and three additional mutants FN6265, FN6162, and FN6142 belonging to independent symbiotic complementation groups (unpublished results). The normalized, filtered, and sorted sequence reads were mapped to the M. truncatula genome assembly (A17r5.0) using the UGENE open‐source tool (Okonechnikov et al., 2012) and the mapped reads were visualized in the Integrated Genome Browser (IGB) (Freese et al., 2016). We identified a region between genetic markers Mtb319 and Chr3_6,7M where no reads were mapped in FN9285, but we did identify nodule‐specific transcripts in the other three mutants, indicating a potential deletion in FN9285 (Figure S2b). A ~250‐kb deletion was confirmed by PCR‐based markers (Figure S2c) that removed 33 genes in FN9285 including 16 nodule‐specific genes (Table S1). Some nodule‐specific genes were not annotated correctly in the M. truncatula genome assembly (A17r5.0) and a more thorough sequence analysis and a closer look at the database of NCRs (de Bang et al., 2017; Montiel et al., 2017) revealed that together with seven already identified NCRs, one hypothetical and one putative protein gene also encode for NCR peptides. Ultimately, 9 NCR genes, NCR132 (3 g0083011), NCR583 (3g0083081), NCR087 (3g0083091), NCR406 (3g0083111), NCR301 (3g0083131), NCR165 (3g0083221), NCR136 (3g0083281), NCR314 (3g0083301) and NCR086 (3g0083311) were identified in the deleted region of FN9285 (Figure 3a).

Figure 3.

The deleted region and gene models within the symbiotic locus of FN9285.

(a) Genetic mapping and transcriptome analysis identified a deletion, spanning nearly 250 kb, in the genome of the symbiotic mutant FN9285. The deletion removed a pile of genes including a cluster of N odule‐specific C ysteine‐R ich (NCR) genes (only NCRs are shown).

(b) Multiple sequence alignment and (c) phylogenetic analysis of the mature peptides encoded by the deleted NCR genes in the symbiotic locus of FN9285 and two previously identified essential NCRs, NCR169 and NCR343.

(d) Sequence alignment of peptides NCR314 and NCR086. Black arrow shows the cleavage site of the predicted signal peptides. Coloration of amino acids is based on Robert Fletterick's ‘Shapely Models’ (http://openrasmol.org/doc/rasmol.html#shapelycolours).

(e) Expression levels of the deleted NCRs were analyzed in wild‐type nodules 14 days post‐inoculation with Sm2011 using RT‐qPCR.

(f) The structure of NCR314 and NCR08 6 genes. Exons are presented by thick black arrows and a thin black line shows the intron.

Since some NCR peptides were found to be essential for the complete bacteroid differentiation and persistence in symbiotic nodules, we hypothesized that lack of one of the NCR genes might be responsible for the symbiotic phenotype of FN9285. To find out which of the nine NCR peptides might be crucial for the effective symbiotic interaction between M. truncatula and Sm2011, further analyses of the NCR genes and the encoded peptides were performed. The sequence alignment of the mature peptides of the 9 NCRs and three crucial peptides, NCR169, NCR343, and NCR‐new35 revealed that all peptides belong to the group of NCRs containing four cysteine residues in conserved positions (Figure 3b). The phylogenetic analysis has revealed that six peptides (NCR301, NCR087, NCR132, NCR136, NCR406, and NCR583) from the identified gene cluster show a closer relationship and NCR165 was positioned closer to the previously identified essential NCRs (Figure 3c). We found that the mature peptides of NCR086 and NCR314 are identical. The alignment of the nucleotide sequences of NCR086 and NCR314 identified four nucleotide differences between the coding sequences of the two genes which result in two amino acid substitutions solely in the signal sequences of the peptides (Figure 3d). We compared the isoelectric points (pI) of the 9 NCRs (Montiel et al., 2017); the charges of the mature peptides of NCR087, NCR301 and NCR132 (pI = 4.19, 4.19, 4.39) are anionic, while the mature products of NCR406 and NCR165 are slightly anionic (pI = 6.48 and 6.22). The charges of the mature peptides of NCR583, NCR136, NCR086, and NCR314 are cationic (pI = 9.4, 8.64, 8.66, and 8.66, respectively). Based on the net charge of the nine mature peptides, we could not pinpoint which NCR gene might be responsible for the symbiotic phenotype of FN9285. To further investigate the nine NCRs in the deleted region, we monitored the relative transcript level of these NCRs in M. truncatula A17 nodules 14 dpi with Sm2011 using RT‐qPCR. All NCRs were clearly expressed in the nodules at various levels (Figure 3e); NCR314 and NCR406 showed relatively low transcript levels compared with the other seven NCRs (NCR132, NCR583, NCR087, NCR301, NCR165, NCR136, and NCR086), whose expression varied between 30 and 600 relative ratios. Due to the lack of a unique feature that would discriminate one NCR gene from the others, all nine NCRs were considered a potential causative gene for the symbiotic phenotype of FN9285.

The lack of NCR086 and NCR314 causes the symbiotic defect of FN9285

To identify the gene or genes responsible for the symbiotic defect in the FN9285 mutant, we introduced the genomic fragments of the genes NCR132, NCR583, NCR087, NCR406, NCR301, NCR165, NCR136, NCR314, and NCR086 into FN9285 using Agrobacterium rhizogenes‐mediated hairy root transformation. The gene constructs contained native promoters ranging from 900 bp up to 1.3 kbp and native 3′ UTRs within the range of 1.2 kbp. We used DsRed and YFP fluorescent protein markers to detect transformed roots. The roots of transformed plants were inoculated with Sm2011 (pXLGD4), and nodule sections were stained for β‐galactosidase activity at 7 wpi. Since no rhizobia‐infected symbiotic cells were detected in the nitrogen fixation zone of white, slightly elongated mutant nodules of FN9285 (Figure 2a,b); therefore, the successful rescue of the symbiotic defect was mainly assessed by the size and color of the nodules and the colonization of ZIII. On the hairy roots of FN9285 transformed with the empty vector or with the genes NCR132, NCR583, NCR087, NCR406, NCR301, NCR165, and NCR136, merely white and undeveloped nodules with rhizobia‐colonized cells only in ZII were formed, indicating that these NCR genes were unable to restore the effective symbiosis in the mutant plant (Figure 4). However, the mature nodules developed on the roots of FN9285 transformed either with the NCR086 or NCR314 construct were all elongated and pink, and the ZIII of these nodules were invaded by rhizobia similarly to the empty vector‐transformed nodules of wild‐type plants. In addition to the formation of the functional nodules, FN9285 plants transformed with the genes NCR086 and NCR314 developed dark green leaves and grew bigger shoots compared with the plants transformed with the empty vector or the other seven NCR genes (Figure 4). We measured the biomass of the aerial part of mutant plants transformed with the 9 NCR genes. The dry weight of FN9285 mutant plants transformed with NCR132, NCR583, NCR087, NCR406, NCR301, NCR165, and NCR136 was similar to that of empty vector‐transformed mutant plants further confirming that these NCR genes were unable to restore the effective symbiotic interaction in mutant FN9285 (Figure S3). In contrast, the dry weight of the green part of FN9285 plants transformed either with NCR086 or NCR314 was comparable with the empty vector‐transformed wild‐type plants suggesting that complementation of FN9285 with NCR086 or NCR314 was effective. The development of functional nodules and the restoration of the symbiotic performance of the transformed FN9285 plants indicated that loss of genes NCR086 and NCR314 was causative of the symbiotic defects of this mutant.

Figure 4.

Both NCR314 and NCR086 restore the effective symbiosis in FN9285 assess based on the growth rate of transformed plant and the morphology and colonization of nodules.

The NCR genes were introduced into FN9285 roots using Agrobacterium rhizogenes‐mediated hairy root transformation. Plants transformed with either an empty vector (EV) or NCR peptide genes were inoculated with Sinorhizobium meliloti 2011 expressing the lacZ marker gene. Longitudinal sections of 7‐week‐old nodules were stained for β‐galactosidase activity. Transgenic roots were identified either by green or DsRed fluorescent protein fluorescence. Nodules on FN9285 roots transformed with NCR314 and NCR086, as well as on wild‐type (WT) roots transformed with EV, exhibited the characteristic zonation of indeterminate nodules, indicating effective symbiosis. Gene constructs of NCR136, NCR165, NCR301, NCR406, NCR087, NCR583, and NCR132 did not restore the effective symbiosis on transformed roots of FN9285, suggesting that these genes are not essential for the symbiotic interaction between Medicago truncatula A17 and S. meliloti 2011. Scale bars: for whole plants—2 cm, nodule images—200 μm, nodule sections 65 μm.

Gene editing confirms the requirement of NCR086 and NCR314 for M. truncatula‐Sm2011 symbiosis

We have recently used the CRISPR/Cas9 gene editing system to assess the requirement of NCR genes for nitrogen‐fixing symbiosis in M. truncatula (Güngör et al., 2023). To prove the necessity of NCR086 and NCR314, two guide RNAs targeting both genes were designed and one of the NCR genes in the deleted region of FN9285, NCR583 was also targeted by a single guide RNA construct (Figure 5a). To induce mutations in genes NCR086, NCR314, and NCR583 by the Cas9 endonuclease, the sgRNA constructs were introduced into the roots of wild‐type M. truncatula A17 genotype using A. rhizogenes‐mediated hairy root transformation. The roots were inoculated with Sm2011 (pXGLD4) and nodules developed on transformed roots, confirmed by the fluorescence of DsRed protein, were harvested and analyzed for gene editing events and assessed for their symbiotic phenotype applying the previously described method (Güngör et al., 2023). Shoot development, nodule morphology, and colonization of transformed plants were analyzed and the amplicon sequences of the targeted region were determined by using a MiSeq platform and analyzed with the online tool CRISPResso2 (Clement et al., 2019). M. truncatula A17 plants carrying induced insertion and deletion mutations in NCR583 retained the effective (Fix+) symbiotic phenotype; they developed dark green leaves and formed elongated pink nodules fully colonized in ZIII detected by SYTO13 staining, very similar to empty vector‐transformed wild‐type plants (Figure 5b,c). The CRISPR/Cas9‐mediated simultaneous knockout of the NCR086 and NCR314 alleles showed an ineffective (Fix‐) symbiotic phenotype; they developed yellowish leaves and formed slightly elongated white nodules. SYTO13 staining and lacZ activity showed that the nitrogen fixation zones of these nodules were not colonized by rhizobia and rhizobia were hardly elongated in the IZ indicating that NCR086 and NCR314‐edited nodules were not functional (Figure 5b). The CRISPR/Cas9‐mediated gene editing of NCR086 and NCR314 proved the requirement of these peptides for the nitrogen‐fixing symbiosis between Sm2011 and M. truncatula and demonstrated that the CRISPR/Cas9 system is a valuable tool for functional analysis of NCR genes in M. truncatula.

Figure 5.

Confirmation of the requirement of NCR086 and NCR314 for the symbiotic interaction between Medicago truncatula and Sinorhizobium meliloti 2011 using CRISPR/Cas9 gene editing.

(a) A single guide RNA (sgRNA) targeting NCR583 and two sgRNAs targeting both NCR086 and NCR314 were designed. The sgRNAs and PAM sequences in red are shown. The constructs were introduced into the roots of M. truncatula A17 using Agrobacterium rhizogenes‐mediated hairy root transformation. The transgenic roots were selected based on DsRed fluorescence.

(b) The phenotypic analysis of gene‐edited plants and nodules. Wild‐type plants transformed with empty vector (EV) or the construct targeting NCR583 showed no signs of nitrogen deficiency and nodules formed on these transformed hairy roots were elongated and pink and the SYTO13‐staining showed the regular colonization of the characteristic zones of indeterminate nodules. Targeted mutagenesis of NCR086 and NCR314 resulted in plant showing the symptoms of nitrogen starvation (retarded growth and yellowish leaves) and white nodules formed on transgenic roots were slightly elongated. The colonization of NCR086/NCR314‐targeted nodules, visualized by SYTO13 and lacZ staining, was impaired in ZIII resembling the nodulation phenotype of FN9285. Bars: 2 cm for plant images and 200 μm for nodule images whole nodule sections—65 μm, enlarged nodule image—20 μm.

(c) The allele sequences of randomly selected nodules edited in genes NCR583, NCR086 and NCR314. The amplicons of the targeted regions were genotyped using Illumina next generations sequencing and sequences were analyzed using the CRISPresso2 tool. The gRNA and PAM (in red) sequences are shown above the allele sequences. Nucleotides in red squares show insertions, dashes indicate deleted nucleotides. The targeted mutations resulted in nonfunctional nodules, confirming the essential roles of NCR314 and NCR086 in symbiosis between M. truncatula and S. meliloti 2011. Functional nodules were found on NCR583‐edited roots, suggesting the nonessential function of NCR583 in this interaction.

The expression levels of NCR086 and NCR314 differ significantly

The genes NCR086 and NCR314 are able to restore the symbiotic defect of FN9285 mutant plants (Figure 4). The coding sequences of NCR086 and NCR314 differ in four nucleotides that results in the change of two residues at the positions 20 and 21 in the signal peptide but the two mature peptides are identical (Figure 3b,d). The extremely close similarity between the NCR086 and NCR314 suggested that these two NCR genes might have evolved recently. We analyzed the gene content of the NCR gene clusters, deleted in FN9285, between the genome sequences of two widely used M. truncatula genotypes, Jemalong A17 and R108 (recently renamed Medicago littoralis; Choi et al., 2022). The analysis of their genome sequence revealed seven NCR genes in R108, which were in the same order and orientation as their Jemalong A17 counterparts (Figure S4a, Table S2), suggesting that the duplication of the NCR086/NCR314 gene and the gain of NCR165 occurred in Jemalong A17 (or the loss of NCR165 in R108) after the split of the two M. truncatula lines. These results were confirmed with the mixed sequence traces of genes NCR086 and NCR314 in Jemalong J5 and A17 at the position of the nucleotides resulting in different residues between the two peptides between R108 and A17 plants (Figure S4b).

We observed differences of several orders of magnitude in the expression levels between NCR086 and NCR314 in native roots (Figure 3e). We scrutinized the expression of NCR086 and NCR314 in nodules at 4, 7, 10 and 14 dpi with Sm2011 by qPCR using primers specific for the 3′ UTRs of NCR086 and NCR314. The expression levels of both genes peak at 7 dpi and maintained at later time points but the transcript level of NCR086 was about two orders of magnitude higher compared with NCR314 (Figure S5). Although NCR314 shows much lower level of expression, it is still able to restore the symbiotic defect of FN9285 (Figure 4). We speculated that the lower expression detected by qPCR is due to the fact that NCR314 is expressed in fewer cell layers or limited to less nodule zones. To analyze the spatial expression of NCR086 and NCR314, the same promoter regions used in the complementation experiments were fused to β‐glucuronidase (GUS) reporter gene and introduced into the roots of M. truncatula wild‐type plants using A. rhizogenes‐mediated hairy root transformation. GUS staining of nodules showed that the promoters of both genes are active at 3 weeks post inoculation (wpi) with Sm2011 in the infection zones (ZII), most robustly in interzones (ZII‐III), and in the fixation zones (ZIII) (Figure 6a). The RT‐qPCR analysis of GUS expression in the transformed roots detected a similar level of expression suggesting that the different expression levels of NCR086 and NCR314 are not due to the different activity of the two promoters but caused by other regulatory elements (Figure 6b).

Figure 6.

The transcriptional activity of NCR314 and NCR086 promoters.

(a) Promoters of NCR314 and NCR086 genes fused with the β‐glucuronidase (GUS) gene were introduced into the roots of WT plant using Agrobacterium rhizogenes‐mediated hairy‐root transformation. Empty vector (EV) transformed roots were used as a control. GUS activity was assessed in nodule sections at 3 weeks postinoculation (wpi) with Sm2011. Scale bars: 200 μm.

(b) Relative expression of the GUS gene in nodules of WT plants transformed with the constructs pMtNCR086::GFP‐GUS and pMtNCR314::GFP‐GUS revealed a similar promoter activity of the two NCRs. Error bars indicate SEM. ns (nonsignificant) indicates statistical significance based on the Tukey test.

To assess how the regulatory elements affect the complementation capacity of NCR086 and NCR314, constructs containing the promoters, the coding sequences and 3′UTRs of NCR086 and NCR314 in different combinations were prepared and introduced into the FN9285 mutant roots using A. rhizogenes‐mediated hairy root transformation. All the combinations of the gene constructs were able to restore the symbiotic phenotype of FN9285 assessed based on the gross phenotype, the biomass of transformed plants, formation of functional pink nodules and the colonization of the fixation zone (Figures S6a,b and S7). Based on these results, we concluded that further experiments are required to identify the regulatory sequence responsible for the lower transcriptional activity of NCR314.

DISCUSSION

Rhizobia in the nodules of IRLC and some Dalbergoid legume species (Aeschynomene sp. and Arachis hypogea) undergo terminal differentiation (Czernic et al., 2015; Oono & Denison, 2010; van de Velde et al., 2010). This irreversible bacteroid transition is mediated by NCR peptides in M. truncatula (van de Velde et al., 2010) and legumes in which irreversible transformation occurs also contain NCR or NCR‐like peptide genes (Czernic et al., 2015; Montiel et al., 2017; Quilbé et al., 2021; Raul et al., 2022). The terminal differentiation of bacteroids in IRLC legumes is associated with the chromosomal multiplication of rhizobia and the arrest of cell division leading to different degrees of cell elongation, which correlates with the number of symbiotic NCR peptides present in each IRLC legume species (Montiel et al., 2017). So far, only M. truncatula symbiotic mutants deficient in NCR peptides have been identified, and the lack of some special NCR has always resulted in incomplete differentiation of bacteroids and the formation of nodules with noncolonized nitrogen fixation zones (Gao et al., 2023; Horváth et al., 2015; Kim et al., 2015; Sankari et al., 2022; Zhang et al., 2023). The symbiotic phenotype of FN9285 was in accordance with these previous findings; FN9285 formed slightly elongated nodules wherein the infection and interzones were invaded by rhizobia. The bacteroid differentiation, measured by the size and the DNA content, was incomplete similarly to two recently identified mutants defective in NCR343 and NCR‐new35 peptides, respectively (Horváth et al., 2023). However, a slightly higher proportion of shorter bacteroids in FN9285 compared with the other two ncr mutants and the colonized region restricted to a narrow distal part of FN9285 nodules might indicate that bacteroid differentiation is blocked earlier in FN9285.

The deletion cosegregating with the symbiotic locus of FN9285 covers a large proportion of one of the symbiotic islands identified based on the transcriptional activity of several colocalized genes expressed in the nodule differentiation zone (Pécrix et al., 2018). The deletion removed 33 genes or gene models including nine NCR genes and an additional seven nodule‐specific genes. Based on the symbiotic phenotype of FN9285, the NCRs were the primary candidates responsible for the symbiotic defect of FN9285. Genetic complementation proved that the loss of two members of the NCR gene cluster, NCR086 and NCR314 encoding the same mature peptide, resulted in the incomplete bacteroid differentiation and the ineffective symbiotic phenotypes of FN9285 (Figure 3d,f). This experiment also indicated that the other NCRs in the cluster are not required for the studied symbiotic interaction, which leaves the question of whether they are required for any symbiotic relationship unanswered. To confirm the requirement of NCR314 or NCR086, the CRISPR/Cas9 gene editing was applied, which also verified that NCR583 is dispensable for this symbiotic interaction and, in addition, demonstrated that a reverse genetic approach can be a feasible strategy in the case where gene identity has to be proven from among many deleted genes. A partially overlapping deletion was identified in the symbiotic M. truncatula mutant FN007, which continues upwards from the end of the deletion (MtrunA17Chr3g0082991) in FN9285 and ends at the gene MtrunA17Chr3g0083261 (Shen et al., 2023). The overlap between the deletions in FN9285 and FN007 contains six NCRs (NCR132, NCR583, NCR087, NCR406, NCR301, and NCR165) and our findings indicate that the lack of any of these NCR genes in FN007 does not cause the defective symbiotic phenotype.

Some other crucial NCR genes are also colocalized with their close counterparts in the M. truncatula genome; NCR211 is located next to its closest homolog NCR178 (Kim et al., 2015), and NCR343 is associated with three other related NCRs (Horváth et al., 2023). Here, we identified a cluster of NCRs, which probably evolved by a series of tandem duplications from an ancestral NCR gene. This hypothesis is supported by the presence of two genes encoding the identical mature peptide in M. truncatula A17 genotype, while only one copy of this gene was found in the M. littoralis R108. The sequence comparison of the A17, J5 and R108 alleles of the coding sequence of NCR086 suggests that the duplication of NCR086 in Jemalong resulted in a novel copy termed NCR314. Interestingly, the copy of NCR086 in R108, termed NC314 in the paper published by Achom et al. (2022), was found showing a circadian rhythm of transcription that peaks at nights. The conserved evening element (EE), the binding site of the circadian clock‐associated transcription factor LATE ELONGATED HYPOCOTYL (LHY), was detected in the promoter of NCR314 ^R108 (Achom et al., 2022), but the sequence analysis of our study revealed EE only in the promoter of NCR086 ^A17 but interestingly not in NCR314 ^A17 . Based on these results, we believe that NCR086 is the more ancient version of the gene, which generated NCR314 through duplication in the A17 genotype (NCR314 ^A17 ), and therefore we believe that the NCR086 gene and not NCR314 is present in R108. In addition to the duplication of NCR086, another difference was the presence of NCR165 in this gene cluster of the A17 genotype.

Generation of new members in gene families through repeated tandem duplications and the neofunctionalization of duplicated genes are very well documented in the case of the plant resistance genes (Leister, 2004; Ronald, 1998). Beyond the genes of the plant and animal immune systems, has also been reported in many processes in plants such as cell wall or storage protein production and photosynthesis (Gibbs et al., 1989; Huo et al., 2018; Monson, 2003; Tao et al., 2023). The generation of an extra copy of a gene produces raw material for novel gene function, which may contribute to the adaptation to new environmental conditions. Resistance (R) genes are often arranged in the plant genome in clusters and the duplication and the rapid evolution of the members in a cluster can lead to new copies conferring resistance to a new pathogen (Ronald, 1998). Although, the exact function of each NCR peptide is not yet known, based on the different composition of the NCR gene cluster identified between the two M. truncatula A17 and M. littoralis, we can assume a similar evolutionary process of NCR genes in the genome of Medicago species that generates diversity and later, functional novelty. The diversification of NCR086 and NCR314 has been probably triggered by gaining new regulatory sequences in NCR314 ^A17 which confers lower transcriptional activity in the M. truncatula nodules. Despite this change, NCR314 ^A17 retained the ability to restore the effective nodulation of FN9285.

The analysis of the temporal and spatial activity of the NCR086 and NCR314 promoters revealed that they are active in the IZ and ZIII of the symbiotic nodules similarly to the previously identified crucial NCR genes, NCR169, NCR211, and NCR343 (Horváth et al., 2015, 2023; Kim et al., 2015). However, we found that the transcriptional activity of NCR314 was about 100 times lower than that of NCR086 but it was still able to restore the effective symbiosis in FN9285. Further studies are required to identify the regulatory sequences which contribute to the expression of NCR314, and potentially other low level expressed NCRs such as NCR‐new35 (Horváth et al., 2023).

The host‐controlled terminal bacteroid differentiation occurs in several lineages of legumes (Oono et al., 2010) and the analysis showed that the symbiotic performance of the terminally differentiated rhizobia strains exceeded that of the reversibly differentiated bacteria (Oono & Denison, 2010). In M. truncatula, six peptides of a ~700‐member NCR gene family have been identified to be crucial for effective symbiosis; however, these studies have been able to associate a specific function only with NCR247 (Gao et al., 2023; Horváth et al., 2015, 2023; Kim et al., 2015; Sankari et al., 2022; Zhang et al., 2023). Here, we added a gene pair of NCR086 and NCR314 to the list of NCR peptides that are essential for complete bacteroid differentiation in M. truncatula nodules. How to identify NCR peptide genes, essential for symbiotic nitrogen fixation, based on the biochemical properties of the encoded peptides or the structure of gene has long been a burning question. The NCR314 and NCR086 genes comprise two exons (like NCR247 and NCR_new35), a prevalent structure of M. truncatula NCR genes but NCR169 and NCR211 consist of three exons while NCR343 composed of a single exon, so the crucial NCRs are encoded by genes with different exon numbers. The expression patterns of this set of NCRs largely overlap indicating their concerted action. However, they have different pIs, varying in a range between 4.78 and 10.15, and no other peptides can complement the loss of any of these, so these NCR peptides are essential. The only feature they have in common is that the mature peptides have four cysteine residues. Except for NCR247, the specific function of the essential NCRs is still elusive, but the cooperative function of these peptides is most likely to optimize the interaction between rhizobia and the host plant. Some of these peptides might even compensate for the negative effects of others (Mergaert, 2018). It has been postulated that a core set of M. truncatula NCRs is required for a complete bacteroid differentiation and/or the loss of their viability outside the host plant (Pan & Wang, 2017; Roy et al., 2020).

MATERIALS AND METHODS

Plant material, rhizobia strains, and growth conditions

Mutant NF‐FN9285 (FN9285) was identified in the fast neutron‐bombarded mutant collection of M. truncatula A17 (Xi et al., 2013). Wild‐type (M. truncatula A17) and mutant plants were germinated and grown as previously described (Domonkos et al., 2017). Briefly, seedlings were grown under symbiotic conditions in a zeolite substrate (Geoproduct Kft., Mád, Hungary) in growth chambers and kept under a photoperiod of 16 h of light and 8 h of darkness at 24°C. Four‐day‐old pre‐grown plants were inoculated with the rhizobia strains Sinorhizobium (Ensifer) medicae WSM419 (WSM419) and ABS7 (ABS7); S. meliloti 1021 (Sm1021) and S. meliloti 2011 (Sm2011) carrying the pXLGD4 plasmid expressing the lacZ reporter gene under the control of the hemA promoter. The inoculation of M. truncatula plants with rhizobia was carried out as previously described (Domonkos et al., 2017).

Acetylene reduction assay

The acetylene reduction assay (ARA) was used to measure the activity of bacterial nitrogenase. Nodules of mutant and wild‐type plants were collected at 5 wpi with strains Sm1021; Sm2011; WSM419 and ABS7. ARA was performed in four biological and three technical replicates and the rate of acetylene reduction to ethylene was calculated using the previously described method (Kovács et al., 2021).

Histological analyses and microscopy

X‐Gal and SYTO13 staining

To assess bacterial colonization, mutant and wild‐type plants were inoculated with rhizobia strains harboring the constitutively active lacZ reporter gene. 14‐day‐old nodules were harvested in 1xPhosphate Buffered Saline (PBS) (pH 7.4), vacuum‐infiltrated and fixed in 4% (wt/vol) paraformaldehyde solution for 3 × 30 sec in an Eppendorf 5301 vacuum concentrator, and then postfixed on ice for 30 min. Fixed nodules were embedded in 5% (wt/vol) agarose before being cut into 65 μm thick longitudinal sections with a Leica VT1200S vibrotome. Nodule sections were stained for β‐galactosidase activity in X‐Gal staining solution for 30 min at 37°C to observe the rhizobium‐infected cells. Stained sections were examined using an Olympus BX41M light microscope with a 10× objective (Olympus Life Science Europa GmbH, Hamburg, Germany), and images were captured with an Olympus E‐10 digital camera.

The symbiotic nodule cell structure and bacteroid morphology were examined on SYTO13 stained nodule sections. Samples were incubated in 5 μM SYTO13 (Thermo Fisher Scientific, Waltham, MA, USA) solution in 1X PBS (pH 7.4) for 20 min at room temperature and imaged using a Leica TCS SP8 confocal laser scanning microscope. The microscope configuration was as follows: Objective lenses: HCX PL FLUOTAR 10× dry (NA: 0.3) and HC PL APO CS2 63× oil (NA: 1.4); zooms: 0.75 and 2.14; excitation: OPSL 488 nm laser (SYTO13, green) and OPSL 552 nm laser (red autofluorescence); spectral emission detector: 510–540 nm (SYTO13) and 558–800 nm (red autofluorescence).

GUS reporter assay

For GUS staining, prefixed nodules were imbibed in 90% acetone for 1 h before sectioning. Whole nodules were vacuum‐infiltrated for 20 min and then stained for 1–4 h at 37°C in a GUS staining solution [2 mM 5‐bromo‐4‐chloro‐3‐indolyl‐d‐glucuronide acid substrate, 2 mM potassium ferricyanide, and 2 mM potassium ferrocyanide in sodium phosphate buffer (pH 7)]. 65 μm thick GUS‐stained sections, prepared similarly as described above, were observed by an Olympus BX41M light microscope.

Lugol staining

To assess the accumulation of starch granules, 65‐μm thick and fixed nodule sections were stained for 15 sec with 0.1 M aqueous potassium iodide solution, then excess staining was removed with 1.5% sodium hypochlorite solution followed by washing with PBS. Lugol‐stained images were taken similarly to the X‐G, Table S2 Gal and GUS‐stained nodules sections.

Measuring the length and DNA content of the bacteroids

The bacteroid length of PI stained bacteroids was measured using ImageJ software on bacteroid images captured by confocal laser scanning microscopy as previously described (Horváth et al., 2023). The DNA content and the size of the bacteroid populations were analyzed by flow cytometry using a CytoFLEX S Flow Cytometer (Beckman Coulter). Bacteroids were prepared for flow cytometry analysis as previously described. In brief, nodules were ground in BEB buffer, pelleted by centrifugation and the supernatant was purified using 50‐μm nylon filter. Heat shocked bacteria suspensions were labeled with 50 μg mL⁻¹ of propidium‐iodide (PI). Following PI staining, 20–30 000 bacteria or bacteroids were measured. The forward scatter, directly proportional to the particle size, was measured at 488 nm wavelength. The PI‐stained DNA was detected at 561 nm excitation wavelength, and the detector applied a 610/20 nm band‐pass filter. The data were plotted using the Kaluza (Beckman Coulter) flow cytometry analysis software. For a more convenient visual evaluation of the histograms, the standard mathematical smoothing process offered by the Kaluza software was performed.

Genetic mapping and identification of deletion in FN9285

Genetic mapping is described in Methods S1. To detect the genetic alteration in the genome of FN9285, RNAseq analyses were carried out on four symbiotic mutants (FN6265, FN6162, FN6142 and FN9285) belonging to independent complementation groups. Nodules of these symbiotic mutants were harvested at 2 weeks post inoculation (wpi) with Sm2011 (pXLGD4) in liquid nitrogen. Total RNA was extracted with TRI Reagent (Sigma) and purified with the Direct‐zol RNA MiniPrep Kit (Zymo Research, Irvine, CA, USA). RNA samples were treated with DNaseI on Zymo‐Spin columns according to the manufacturer's instructions to remove the genomic DNA. RNA sequencing (RNAseq) (quality control of total RNA, library preparation, sequencing) was carried out by the SeqOmics Ltd. (Mórahalom, Hungary) as previously described (Horváth et al., 2023). Trimmed and sorted reads were mapped to mRNA reference sequence subsets obtained from the M. truncatula A17r5.0 genome assembly (MtrunA17r5.0‐ANR‐EGN‐r1.6.cds) (https://medicago.toulouse.inra.fr/MtrunA17r5.0‐ANR/) using the UGENE open source tool (Okonechnikov et al., 2012) and the mapped reads were visualized in the Integrated Genome Browser (Freese et al., 2016). The number of reads of the symbiotic mutants FN6265, FN6162, FN6142 and FN9285 was analyzed in the symbiotic region of FN9285, as determined by genetic mapping. Primers used to validate the deletion in the FN9285 mutant are listed in Table S3.

Gene expression analysis

Gene expression analyses were carried out with quantitative PCR following reverse transcription (RT‐qPCR). Nodules were harvested at different time points described in each experiment. For the expression analysis, total RNA samples were extracted similarly as for RNAseq analysis. Total RNA was quantified on a Nanodrop‐1000 spectrophotometer, and cDNA was prepared from 1 μg total RNA with Maxima Reverse Transcriptase (Thermo Fisher Scientific, Waltham, MA, USA) using oligo‐dT primers according to the manufacturer's protocol in a final volume of 20 μL. qPCRs were performed with the LightCycler 96 system (Roche Holding AG. Basel, Switzerland) using qPCRBIO SyGreen Mix (PCR Biosystems, London, UK) in a total reaction volume of 12 μL. The qPCRs were performed on cDNA samples generated from three independent biological repetitions. A housekeeping gene with a ubiquitin domain (MtrunA17_Chr3g0126781) was used as a reference, and data were analyzed using the LightCycler 96 SW1.1 software. Primers used for qRT‐PCR are listed in Table S3.

Molecular cloning

All constructs were generated in pKGG_RR and pKNGG_YR cloning vectors using the Golden Gate cloning system. The process of creating the pKGG_RR and pKNGG_YR vectors and the generation of the constructs to exchange the different parts of genes NCR314 and NCR086 are detailed in Methods S2. For all complementation constructs, DNA fragments containing 900–1900 bp long regions upstream of the translation start site of each NCR gene, the genomic fragments of the NCR genes and their 3′ UTRs (490–1200 bp long region downstream of the translation stop site), respectively, were amplified with the Phusion DNA Polymerase enzyme (Thermo Fisher Scientific, Waltham, MA, USA) using the BAC clones mth2‐10A20 and mth2‐17J20 as template DNA and the NCR gene‐specific primers (Table S3).

The activity of the NCR314 and NCR086 genes was analyzed with promoter‐GUS reporter assay. For these experiments, constructs containing the same promoter sequences of NCR314 and NCR086 as used for complementation and the PCR amplified GFP‐GUS gene were assembled in the pKGG_RR vector.

Target gene selection and generation of CRISPR/Cas9 gene editing constructs

sgRNAs for NCR583 and NCR086/NCR314 were designed using the online tool http://crispor.tefor.net/ (Concordet & Haeussler, 2018) set for the M. truncatula A17 (MtrunA17r5.0) genome (Pécrix et al., 2018). In each case, the second exon of the gene was targeted. To minimize the likelihood of off‐target alterations, all selected sgRNAs had at least 2–3 mismatches compared to the potential off‐target sites. Targets with the GN19 NGG motif were selected to maximize expression of the MtU6.6 promoter (Güngör et al., 2023). Plasmid constructs were generated in the vector pKSE466_RR (Güngör et al., 2023) for gene editing according to the recommendations described previously (Xing et al., 2014). For NGS amplicon sequencing, appr. 200‐bp genomic regions, including the target site, were amplified with the Phusion DNA polymerase enzyme (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's guidelines. Samples were sequenced using an Illumina platform and at least 10 000 reads/sample were generated. The NGS sequence analysis was carried out by utilizing the CRISPresso2 software (Clement et al., 2019). The primers used for genotyping of gene editing events are listed in Table S3.

Hairy‐root transformation

All constructs were introduced into the mutant or wild‐type plants using A. rhizogenes‐mediated hairy root transformation (Boisson‐Dernier et al., 2001). The transgenic roots were selected based on the DsRed or YFP fluorescence using a Leica MZ10F fluorescence stereomicroscope. Primers used for generating constructs are listed in Table S3. Transformation experiments were repeated 3–4 times with a minimum of 40 plants/constructs.

Protein sequence analysis

Multiple sequence alignments and phylogeny analyses of NCR peptide sequences were created by using the CLC Genomics Workbench 9.5.3 program with default settings (Horváth et al., 2023). For multiple peptide alignment, the “accurate alignment” algorithm of CLC was used (parameters: gap open cost:10; Gap extension cost: 1.0; End gap cost: free) while the phylogenetic tree was generated using the UPGMA algorithm (parameters: Distance measure: Kimura Protein; Bootstrap: 100 Replicates).

AUTHOR CONTRIBUTIONS

PK, and ÁD conceived and designed the project. YC and RC carried out the mutant screen and provided genomic information. FS, JBB, BH, SK, MK, and DL performed the experiments, ÁD and CV carried out the flow cytometry analysis. KL and GR conducted the acetylene reduction measurements. ÁD, PK, and FS wrote and edited the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors report no conflicts of interest.

Supporting information

Figure S1. Starch distribution and bacteroid morphology in ZIII are defective in FN9285 nodules. (a) Starch granules, visualized by Lugol's iodine staining, are accumulated in the symbiotic cells in the interzone and to a lesser extent in ZIII of WT nodules. Starch accumulation was detected in the extended interzone of mutant nodules and the ZIII cells were devoid of starch granules, suggesting a block in metabolic transition in FN9285 nodules. (b) White light confocal images show accumulation of starch granules in ZII and IZ of mutant nodules; (c) White light and fluorescent confocal images show accumulation of starch granules in ZIII of mutant nodules; Confocal images Bars: 200 μm for whole nodules; 20 μm for enlarged images. ZII, infection zone, IZ, interzone; IZ‐ZIII, transition between IZ and ZIII (nitrogen fixation zone), ZIIId and ZIIIp, distal and proximal part of nitrogen fixation zone, respectively. *: starch granules; b, bacteroids; it, infection thread; n, nucleus.

Figure S2. The identification of the genomic region deleted in symbiotic mutant FN9285. (a) The symbiotic locus of FN9285 was identified by genetic mapping on chromosome 3 (LG 3) of Medicago truncatula between the genetic markers Mtb319 and Chr6,7 M. (b) A deletion in the FN9285 genome was identified by mapping RNAseq reads of FN9285, FN6265, FN6162, and FN6142 deletion mutants to the genomic sequence of M. truncatula A17 between genetic markers Mtb319 and Chr_6,7M in LG 3 using the Integrated Genome Browser. (c) The confirmation of the deleted region in FN9285 mutant was carried out using PCR‐based markers developed for gene models in the region. The PCR amplification was carried out using the total DNA of a heterozygous plant from the mapping population (1), the mutant FN9285 (2), and the A20, the wild‐type parent of the mapping population (3). Gene IDs of the tested genes are shown below the gel images.

Figure S3. Genetic complementation of FN9285 with genes NCR314 and NCR086 restores the plant gross of NF9285. Shoot dry weight of transformed plants shows that introducing NCR314 or NCR086 genes into FN9285 roots, using the Agrobacterium rhizogenes‐mediated hairy‐root transformation leads to a significant increase in the plant gross under symbiotic conditions, indicating the functional symbiosis. The transformation with the genes NCR087, NCR136, NCR132, NCR583, NCR165, NCR301, and NCR406 did not complement the reduced plant weight of FN9285. The values represent the mean + standard error of the mean (SEM) of at least 23 plants, and asterisks (**) highlight statistical significance (P < 0.01), ns—nonsignificant determined by the Tukey test.

Figure S4. Comparison of the gene content of Medicago truncatula genotypes A17 and Medicago littoralis R108 in the deleted symbiotic region of FN9285 on chromosome 3. (a) The orientation of genes and gene models is predicted based on the annotation of M. truncatula A17 r5.0 genome portal (https://medicago.toulouse.inra.fr/MtrunA17r5.0‐ANR/) and the genome assembly of R108 (Genbank Bioproject RJNA368719). NCR genes are highlighted with red arrows, other genes and the border markers are represented by blue and gold arrows, respectively. The designation “A” indicates the genes in the A17 ecotype, while “R” represents the R108 genotype. The gene numbers show the genes listed in Table S2. (b) Sanger sequencing of PCR amplified fragments of the coding sequence of the NCR086/NCR314 gene pair shows the mixed sequences (polymorphisms) of NCR086 and NC314 in M. truncatula A17 and in Jemalong (J5), the parental ecotype of A17 accession, which indicates a presence of the duplication of the gene pair in these plants. No polymorphisms were detected in R108, indicating the existence of NCR086/NCR314 in a single copy in R108.

Figure S5. Expression analysis of NCR314 and NCR086 genes during nodule development. The relative expression of NCR086 (a) and NCR314 (b) was analyzed by reverse transcription quantitative polymerase chain reaction (RT‐qPCR) in wild‐type (WT) nodules at 4, 7, 10, and 14 days postinoculation (dpi) with Sm2011. Error bars indicate SEM.

Figure S6. The effect of the different elements of genes NCR314 and NCR086 on the gross phenotype of mutant FN9285. (a) All the gene constructs, containing the promoter, the coding, and terminator sequences of NCR314 and NCR086 in different combinations, transformed into the roots of FN9285 restored the vigor of mutant plants. (b) The shoot dry weight of the transformed plants with the gene constructs suggests the successful rescue of the reduced gross phenotype of FN9285. Values are presented as mean + SE of the mean of at least 15 plants. Asterisks (**) indicate statistical significance (P < 0.01) based on the Tukey test. (scale bar—2 cm). Blue colored boxes represent elements of NCR314, and orange boxes represent elements of NCR086 in the constructs. p314 and p086: promoter regions including the 5′ UTR of NCR086 and NCR314, respectively; g314 and g086: genomic fragments from translation START and STOP codons of NCR086 and NCR314; t314 and t086: 1200–1400 bp long terminal sequences downstream of the translation STOP codons of NCR086 and NCR314.

Figure S7. Dissecting the activity of the promoters and terminator sequences of NCR086 and NCR314. Nodulation phenotype of FN9285 roots transformed with empty vector (EV) or the gene constructs of different elements of the NCR086 or NCR314 genes. The constructs were introduced into the roots of FN9285 using Agrobacterium rhizogenes‐mediated hairy root transformation. The transgenic roots were identified based on DsRed fluorescence. Longitudinal nodule sections were stained for β‐galactosidase activity 7 weeks postinoculation (wpi) with Sm2011. All nodules on the transformed roots of FN9285 exhibited wild‐type morphology and rhizobia colonization similarly to the nodules of wild‐type plants transformed with the EV. p314 and p086: promoter regions, including the 5′ UTR of NCR086 and NCR314, respectively; g314 and g086: genomic fragments from translation START and STOP codons of NCR086 and NCR314; t314 and t086: 1200–1400 bp long terminal sequences downstream of the translation STOP codons of NCR086 and NCR314.

Table S1. Genes and gene models in the symbiotic region of FN9285. Gene IDs of border genes are highlighted in green. IDs in pink stand for nodule‐specific cysteine‐rich peptide genes. Red characters indicate noncorrect annotation and white letters show new annotation done manually. Gene IDs in bold show the genes in the overlapping deletion identified in the symbiotic mutant FN007 (Shen et al., 2023). nt—not tested by PCR.

Table S2. Genes shown in the synteny analysis of the symbiotic region of FN9285 between Medicago truncatula A17 and Medicago littoralis R108. Gene IDs of border genes are highlighted in green. IDs in pink stand for nodule‐specific cysteine‐rich peptide genes.

Table S3. Primers used in this study.

Data S1. Genetic mapping.

Methods S2. Generating cloning vectors and constructs.

DATA AVAILABILITY STATEMENT

All data supporting the findings of this study are available with in the paper and within its supplementary materials published online. Plant materials and other data used in this study are available from the corresponding authors (PK and ÁD) upon request.