Medicago truncatula NIN is essential for rhizobial-independent nodule organogenesis induced by autoactive calcium/calmodulin-dependent protein kinase

Abstract

The symbiotic association between legumes and nitrogen-fixing bacteria collectively known as rhizobia results in the formation of a unique plant root organ called the nodule. This process is initiated following the perception of rhizobial nodulation factors by the host plant. Nod factor (NF)-stimulated plant responses, including nodulation-specific gene expression, is mediated by the NF signaling pathway. Plant mutants in this pathway are unable to nodulate. We describe here the cloning and characterization of two mutant alleles of the Medicago truncatula ortholog of the Lotus japonicus and pea (Pisum sativum) NIN gene. The Mtnin mutants undergo excessive root hair curling but are impaired in infection and fail to form nodules following inoculation with Sinorhizobium meliloti. Our investigation of early NF-induced gene expression using the reporter fusion ENOD11::GUS in the Mtnin-1 mutant demonstrates that MtNIN is not essential for early NF signaling but may negatively regulate the spatial pattern of ENOD11 expression. It was recently shown that an autoactive form of a nodulation-specific calcium/calmodulin-dependent protein kinase is sufficient to induce nodule organogenesis in the absence of rhizobia. We show here that MtNIN is essential for autoactive calcium/calmodulin-dependent protein kinase-induced nodule organogenesis. The non-nodulating hcl mutant has a similar phenotype to Mtnin, but we demonstrate that HCL is not required in this process. Based on our data, we suggest that MtNIN functions downstream of the early NF signaling pathway to coordinate and regulate the correct temporal and spatial formation of root nodules.

Figure 1.

Wild type and the non-nodulating mutant phenotypes of 12S and Tnt148. All images are of plants cultivated in vitro with S. meliloti 1021 pXLGD4 for 14 d. A, Wild-type nodulating root on the right and the non-nodulating 12S root on the left. B, Spot inoculation of a wild-type root produces cortical cell division leading to nodule formation. C, Spot inoculation of 12S does not induce cortical cell divisions. The black mark indicates site of S. meliloti application. D, The formation of a shepherd's crook and an infection focus in the center of the crook in wild type. E, Root hairs of 12S demonstrating the characteristic nin phenotype. Arrows indicate root hairs undergoing excessive curling. F, A root hair of Tnt148 that has formed an infection focus and a limited infection thread. These infection events are infrequent in Tnt148 and absent in 12S. Scale bars = 10 mm (A), 1 mm (B and C), 40 μm (D and E), and 20 μm (F).

Figure 2.

Identification of a Tnt1 insertion in the MtNIN locus and the structure of the conserved NIN domains I to VI. A, Tnt1 insertion in the MtNIN locus (Mtnin-2∷Tnt1) cosegregates with the non-nodulating phenotype of Mtnin-2. Seven F2 Nod⁻ or Nod⁺ plants from a backcross between Mtnin-2 and R108 were used to PCR amplify wild-type MtNIN or the Mtnin-2∷Tnt1 insertion. The absence of amplification products using MtNIN-specific primers indicates that all Nod⁻ individuals are homozygous for the Mtnin-2∷Tnt1 insertion. This is confirmed by Mtnin-2∷Tnt1-specific amplification products in the same samples. Nod⁺ individuals are either heterozygous for the Mtnin-2∷Tnt1 insertion or lack it entirely. B, Six highly homologous domains in the NIN primary sequence are depicted schematically showing their relative positions in M. truncatula, pea, and L. japonicus. The three domains of unknown function (I–III) are shown in dark gray. The predicted hydrophobic/transmembrane domain is represented by a white box. The region containing the putative DNA binding and dimerization domain RWP-RK (RWP) is shown in light gray. Black boxes at the C terminus represent the PB1 domains likely to mediate protein-protein interactions. Numbers above pea and L. japonicus domains indicate percent identity compared to M. truncatula. The star above RWP domain of M. truncatula indicates the approximate position of the 11-bp deletion in Mtnin-1. The Mtnin-2 allele contains a transposon insertion immediately upstream of the ATG (not shown). Scale bar indicates 50 amino acids (aa).

Figure 3.

NIN expression requires the early NF signaling pathway. A, In wild-type plants, steady-state transcript levels of MtNIN increased from 2 d to 10 d after inoculation with S. meliloti. In Mtnin-2, little or no transcript was detectable during the same time course. The barely detectable increase in Mtnin-2 is consistent with the weaker Nod⁻ phenotype in this allele. B, The steady-state expression levels of NIN, ENOD11, and RIP1 were compared between mock-inoculated and S. meliloti 1021-inoculated wild type and the indicated mutants using the Affymetrix M. truncatula oligonucleotide microarray. Each square represents the log² fold change of inoculated versus uninoculated samples in triplicate, 1 d postinoculation. Expression levels above zero are depicted in magenta, and expression levels below zero are depicted in green. The non-nodulating mutants nfp, dmi1, dmi2, dmi3, nsp1, and nsp2 are defective in NF signaling, resulting in the abolishment of NIN, ENOD11, and RIP1 expression compared to wild type and hcl in which these genes are expressed normally. The expression profiles of ENOD11 and RIP1 were previously published (Mitra et al., 2004b).

Figure 4.

NF-induced calcium spiking is normal in Mtnin-1. Cytosolic free calcium levels were monitored over approximately 30 min in wild-type and Mtnin-1 root hairs following 1-nm NF treatment (black vertical bar). Both traces are from single representative root hairs stimulated with NF. Inset table summarizes the total number of root hairs treated compared to the number that generated the spiking response.

Figure 5.

NF-induced ENOD11∷GUS expression in Mtnin-1. A, Seven-day-old primary roots of wild type or Mtnin-1 containing ENOD11∷GUS were continuously treated for the indicated times with 1 nm NF followed by GUS staining. Black brackets define the expanded Mtnin-1 GUS staining pattern in Mtnin-1 compared to wild type. Black arrows indicate the approximate center of the NF responsive zone stimulated by the initial application of NF at the start of the experiment. Due to root growth, this zone is outside of the frame of the image at the 4-d time point. B, seedlings were grown for 7 d on media lacking (−N) or containing (+N) nitrate and treated for 12 h with 1 pm NF, followed by staining for GUS activity. Scale bar = 10 mm.

Figure 6.

MtNIN is required for nodule organogenesis induced by autoactive CCaMK. A, Rhizobia induced nodules in wild-type transformed roots. B, Spontaneous nodules induced by the autoactive CCaMK (DMI3^1–311) construct are present on hcl mutants but absent in Mtnin-1 (C). Wild-type roots (A) were inoculated with S. meliloti, but mutants (B and C) were grown in the absence of S. meliloti. Scale bar = 1 mm.

Figure 7.

Model relating NIN to NF perception and downstream events. The autoactivation of nodule organogenesis by CCaMK (DMI3) is dependent on NIN. Calcium spiking and the induction of ENOD11 expression are normal in Mtnin-1, demonstrating that NIN functions downstream of the early NF signaling pathway (indicated by bracket). All of the components of the early NF signaling pathway are also required for the induction of NIN expression. In addition, the expression of ENOD11 outside of the root responsive zone is suppressed by NIN. Recently published data demonstrates that nodule organogenesis requires cytokinin signaling in the root cortex (shown in gray), which is dependent on both NIN and NSP2. The NIN gene is also required for bacterial infection (as is HCL), suggesting NIN is involved in both epidermal and cortical cell responses. The precise relationship between HCL and NIN cannot be resolved in the work presented here. However, given that HCL expression does not depend on the early NF signaling pathway and autoactive CCaMK induces nodules to form in the hcl mutant, we conclude that HCL is not downstream of early NF signaling leading to nodule organogenesis. Arrows do not imply direct interactions.