Evolution of nitrogen-fixing root nodules : Analysis of conserved signalling modules in legumes and Parasponia

Nitrogen represents one of the most important elements for plant growth. Therefore various plant lineages have established a symbiotic relationship with nitrogen-fixing bacteria. One of the hallmarks of plant-microbe symbioses is the nitrogen-fixing nodule endosymbiosis. In this interaction bacteria are housed intracellularly in so-called nodules; specialized organs formed on the plant root or stems. Inside the nodules plants provide the optimal conditions for the bacteria to convert atmospheric nitrogen into ammonia, which they provide to the plants in exchange for photosynthates. Only about 2.5% of the angiosperm families is able to form a nitrogen-fixing nodule endosymbiosis. The over application of chemical fertilizer in agriculture leads to major environmental problems in terrestrial and aquatic ecosystems. Excessive nitrogen deposition causes a loss of biodiversity in natural habitats and nitrogen leaching into the surface waters causes excessive algal blooms. However, without the use of chemical fertilizers the food demands in the world as we know it would not be attainable. Therefore a major aim of the scientific community considers it a major aim to engineer a form of Nitrogen-fixing endosymbiosis in major crop plants, such as rice, wheat or maize.The nitrogen-fixing endosymbiosis with rhizobia or filamentous actinobacteria of the genus Frankia only occurs in four related taxonomic orders. The Fabales, Fagales, Rosales and the Cucurbitales. Collectively called the Nitrogen Fixation Clade (NFC) Within this clade nodulation is restricted to 10 families scattered among mostly non nodulating families. A nitrogen-fixing endosymbiosis can occur with three different types of bacteria: i. Filamentous Actinobacteria form the genus Frankia, nodulating a paraphyletic assembly of 25 genera distributed of 8 taxonomic families. ii. Rhizobia a paraphyletic group of - α, β and ƴ-Proteobacteria, nodulating only two families, the Legumes (Fabaceae) and Parasponia (Cannabaceae). In order to explain this distribution a hypothesis, is the existence of a precursor state or “predisposition” for nodulation. In this way in the first common ancestor of the NFC, an innovation happened, which made it more likely for its descendants to evolve nodulation. Most of the knowledge on rhizobium symbiosis comes from model legumes Medicago truncatula and Lotus japonicus. In these lineages it was discovered that Nodulation is initiated by the perception of Lipo-chitooligosaccharides (LCOs), which are structurally similar to the signals produced by Arbuscular Mycorrhizal fungi(AM). These obligate biotrophic fungi colonize the roots of most land plants, where they exchange nutrients for photosynthates. Besides the similarity of the signalling molecule, many of the downstream signalling components between AM-symbiosis and rhizobium symbiosis are shared. Therefore the signalling pathway is called the Common Symbiosis Signalling Pathway (CSSP). Many common aspects of symbiosis signalling were reviewed in Chapter 2 in more detail.The genus Parasponia, represents the only lineage to nodulate with rhizobium outside of the legume family. It was previously estimated to have gained its symbiosis relatively recently, given its close relationship to the non-nodulating Trema lineage. In order to find this evolutionary precursor state and the innovations that would have initiated a symbiosis with rhizobia. We set out to compare nodulator Parasponia with non nodulating relative Trema in Chapter 3. In contrast to our initial expectations with did not find any gene gains that would correlate to the nodulation trait. Rather we found a pattern of gene loss in close relatives of the nitrogen-fixing Parasponia in the Rosales lineage. In addition we discovered a large overlap in the nodule enhanced gene set of model legume Medicago and Parasponia.  Three genes were found to be consistently lost in close relatives of Parasponia, transcription factor Nodule Inception(NIN), LCO-receptor Nod Factor Perception (NFP2) and a protein related to infection Rhizobial Polar Growth (RPG). In legumes these three genes are essential for rhizobial infection and nodule formation.In Chapter 4, I continued work on the gene family of Lysin-Motif receptor like kinases (Lysm-RKs) in Parasponia, which includes the putative LCO-receptor NFP2. In legumes it was discovered that the LYR-type Nod factor receptor MtNFP/LjNFR5, functions as a heterodimer with a LYK type receptor MtLYK1/LjNFR1. Here I discovered that Parasponia uses at least four LysM-RKs for rhizobial Nod-factor recognition. Of these four receptors, two are of the LYK-type and have intact kinases with phosphorylating ability. These are named PanLYK1 and PanLYK3. These receptors evolved upon an ancient duplication in the eudicots. In addition to a role in rhizobium symbiosis, PanLYK3 is also involved in chitin triggered immunity, indicating a dual functionality for this receptor. The second receptor PanLYK1 has no major phenotype as a single mutant. However only a double panlyk1-panlyk3 mutant can complete block nodule formation and Arbuscular Mycorrhizal infection in Parasponia. This indicates that there seems to be functional overlap but also a distinction between the two LYK-I type receptors.The other two receptors, PanNFP1 and PanNFP2, represent two LYR-type LysM-RKs, with inactive kinase conformations. The duplication which gave rise to these two copies of LYR-I type receptors happened in an ancestor of the NFC. The presence of an NFP-II-type ortholog strictly correlates to the presence of nodulation. The loss of the NFP2 copy in non-nodulating lineages indicates that this receptor is committed to functioning as a stringent LCO-perception protein in symbiotic context. While pannfp2 mutants cannot be infected by Rhizobia or form nodules, they have no apparent phenotype in AM-symbiosis, which further supports its specialized role. PanNFP1 most likely has a role in AM-symbioses as was suggested previously, which would be in line with its presence in most non-nodulating lineages in the NFC-clade. However, besides a minor phenotype in Nodulation a functional role for this receptor in AM-colonization could not be supported.Besides the loss of LCO-receptor NFP2. Orthologues of two other receptors, PanLEK1 and PanCRK11, were consistently lost in the Trema lineage. In Chapter 5, I characterized their respective gene families and show that PanLEK1 plays an essential role in regulating defense responses in the Nodule. Mutants in panlek1 show a reduction in nodule number and a accumulation of phenolic compounds in the nodule. PanCRK11 belongs to a large gene family of Cysteine-Rich Receptor like kinases (CRKs), many of which are regulated in symbiotic context. A genomic cluster of 20 CRKs was targeted by CRISPR-CAS9, which resulted in a reduction of nodule number and infection level in the mutant. These results hint towards a role for cysteine richt kinase receptors in regulating infection thread progression.Most of the genes used the context of Nodulation exist outside of the NFC, where they have a different function. The recruitment of these genes in nodulation required novel Cis-regulatory elements in their respective promoters. Finding these Cis-regulatory elements may prove essential for future engineering efforts in crop species, since they allow the correct spatio-temporal gene expression. NIN-represents one of the most central transcription factors in Nodulation, however its regulation in both Symbiotic infection and Nodule formation was to date not well understood. In Chapter 6, we discovered a novel a Cis-regulatory element required to initiate NIN-expression in the pericycle. This element, proved to be essential for functional complementation of a nin mutant. The expression of this element is in part regulated by cytokinin, which by itself is capable of stimulating nodule organogenesis.The findings in Chapter 3,4 and 5 are not in line with a independent origin of Nodulation, rather they suggest a Single origin of Nodulation in the NFC. This scenario would imply a widespread loss of the Nodulation trait. In the general discussion Chapter 7, I discuss the potential drivers which could have led to the widespread loss of the nitrogen fixing endosymbiosis. I discuss the implications of these findings for the potential of engineering the nitrogen fixation trait in crop species.